Mechanism of Cancer Chemotaxis Defect

May 2, 2012

When we were studying the chemotaxis defect known to exist in leukocytes from cancer patients, we looked at formylpeptide receptor dynamics in considerable detail.  We examined receptor binding, down-regulation, recovery, and receptor-mediated pinocytosis using rabbit peritoneal PMN, human peripheral PMN, and human peripheral monocytes.  The cells were either obtained directly from cancer patients or were from normal subjects.  Cells from normal subjects were pretreated with serum that had been obtained from cancer or normal patients.

The full paper and link to the PDF are shown below.

 Mechanism of Cancer Chemotaxis Defect

 

Mechanism of the Cancer-Related Leukocyte Chemotaxis Defect: 

Formylpeptide Receptor Modulation and Pinocytosis

Amelia H. Janeczek, PhD^1,3, Pierson J. Van Alten, PhD¹, Hernan M. Reyes, MD²,

and Robert J. Walter, PhD²

¹  Department of Anatomy and Cell Biology, University of Illinois at Chicago, Chicago, IL  60612

²  Department of Surgery, Cook County Hospital, Hektoen Institute for Medical Research, Chicago, IL  60612

³  Present address:  Department of Biochemistry, School of Medicine, Boston University, Boston, MA  02118-2394

Address all correspondence to:

Robert J. Walter, PhD

Department of Surgery, Room 905

Hektoen Institute for Medical Research

625 South Wood Street

Chicago, IL  60612

Telephone:        (312) 633-7237/8717           FAX:       (312) 738-3102

Running Title:  Mechanism of Cancer Chemotaxis Defect

Keywords:  formylpeptide, receptors, cancer, chemotaxis, endocytosis

SUMMARY

Monocyte chemotaxis is severely depressed in patients with advanced tumors but the cellular basis for this chemotactic defect is not known.  Pretreatment of normal human leukocytes or rabbit peritoneal neutrophils (PMN) with serum from cancer (CA) patients inhibits both monocyte and PMN chemotaxis as compared to leukocytes pretreated with serum from healthy (CT) blood donors.  Using purified fresh CA patient leukocytes or CA serum-treated normal leukocytes, formylpeptide receptor binding and modulation were quantified using radiolabeled formylmethionyl-leucyl-phenylalanine (³H-FMLP).  The cell surface binding of ³H-FMLP at 4°C was significantly reduced in CA serum pretreated rabbit peritoneal PMN, but not in CA serum pretreated human peripheral blood PMN or mononuclear leukocytes as compared to CT serum pretreated cells.  However, in purified mononuclear leukocytes isolated directly from tumor patients, formylpeptide binding was significantly reduced as compared to those of normal subjects.  PMN from tumor patients exhibited no significant difference in this regard as compared to PMN from control subjects.  The time course and dose response curves observed during formylpeptide receptor down-regulation were similar for CT and CA serum pretreated cells.  Subsequent to down-regulation, the recovery of cell surface formylpeptide binding at 37°C was similar in its rate and extent in CT and CA serum pre­treated cells.  However, significant reductions in ³H-FMLP uptake at 37°C were seen in tumor patient PMN, in CA serum pretreated rabbit PMN, and in CA serum pretreated human PMN and mononuclear leukocytes as compared to controls.  Such reductions in initial formylpeptide cell surface binding and in formylpeptide endocytosis may contribute directly to the cancer-associated depression of leukocyte chemotaxis.

INTRODUCTION

      Monocyte and macrophage chemotaxis is severely impaired in patients with advanced tumors, but neutrophil (PMN) locomotion is unaffected (1,2).  This defect in monocyte chemotaxis may have life-threatening consequences since host leukocytes may be unable to adequately repress bacterial and fungal infections or tumor growth.  Implantation of tumor cells or injection of tumor sonicates into mice results in reduced accumulation of macrophages in response to inflammatory stimuli, as well as decreased resistance to bacterial infec­tion (3,4).  Similarly, treatment of normal leukocytes with serum from patients with advanced tumors, conditioned media from tumor cell lines, as­cites fluid, plasma, or urine from tumor-bearing mice suppresses monocyte (and often PMN) polarization and chemotaxis (2,5‑9).  This chemo­taxis defect is evident in animals and patients bearing any of a large number of tumors and is seen in response to several quite different classes of che­moattractants (for review see 6,10,11).  Many studies have shown that monocyte chemo­taxis returns to normal after the surgical removal of tumor or after treatment with chemotherapy or radiotherapy (12,13).  Taken together, this suggests that tumors may be producing or causing the host to produce an inhibitor of leukocyte chemotaxis (7,14). 

A soluble inhibitor of monocyte chemotaxis, rendered inactive by treat­ment with antibody directed against the retroviral envelope protein p15E, has been found in tumor patient effusions (15,16) and in serum from patients with head and neck cancer (CA) (17).  This inhibitor or its synthetic ana­logues may cause decreased monocyte polarization in response to chemoattract­ant (8,16), alterations in formylpeptide receptor expression (18,19), suppression of the respiratory burst (20), and inhibition of protein kinase C-related cell functions (21).  It was hypothesized in the present study that the chemotactic defect in tumor patient monocytes or in CA serum-treated leukocytes resulted from alterations in reception or transduction of the signal for chemotaxis.  To examine this possibility, formylpeptide chemoattractant receptor binding was examined in leuko­cytes isolated from tumor patients, in CT or CA serum pretreated human leuko­cytes isolated from normal subjects, and in CT or CA serum pretreated rabbit peritoneal PMN. 

           

Significant reductions in formylpeptide binding were observed in purified tumor patient mononuclear leukocytes as well as in CA serum pretreated rabbit peritoneal PMN.  Chemoattractant uptake at 37°C was significantly reduced in CA serum pretreated rabbit PMN, human PMN, human mononuclear leukocytes, and in tumor patient PMN.  These alterations in surface binding and internalization of formylpeptide may contribute to the depression of chemotaxis exhibited by these cells. 

Abbreviations used: 

BSA, bovine serum albumin; CA, cancer; CT, control; DMSO, dimethylsulfoxide; EDTA, edetic acid; FMLP, N-formyl-methionyl-leucyl-phenylalanine; HBS, HEPES buffered salt solution; HBSS, Hanks' balanced salt solution; HEPES, N-2-hydroxyethylpiperazine-N-2'-ethanesulfonic acid; MLCK, myosin light chain kinase; PMN, polymorphonuclear leukocyte.

                                               METHODS AND MATERIALS

Buffers

For most of the studies described here, HEPES-buffered saline (HBS) containing 140 mM NaCl, 10 mM KCl, 10 mM N-2-hydroxyethylpiperazine N-2-eth­anesulfonic acid (HEPES), 5 mM glucose, and 2 mg/ml bovine serum albumin (BSA), pH 7.4 was used.  For rabbit peritoneal PMN, Hanks' balanced salt solution (HBSS) containing 10 mM HEPES, 5 mM glucose, 136 mM NaCl, 5 mM KCl, 373 nM Na₂HPO₄, 734 nM KH₂PO₄, pH 7.2 was used. 

Rabbit Peritoneal Neutrophils

Glycogen-elicited peritoneal PMN were collected from rabbits in heparin-containing tubes on ice, centrifuged at 500 x g for 10 minutes, and washed in HBSS containing 2 mM EDTA.  Contaminating erythrocytes were removed by brief hypotonic or ammonium chloride lysis, cells were washed in HBSS supplemented with 0.2 mM CaCl₂ and 1 mM MgSO₄, and kept on ice until use.  These cell preparations consisted of >95% PMN.

Isolation and Purification of Human Leukocytes

Peripheral venous blood samples were obtained with informed consent from healthy adult volunteers and from patients admitted to Cook County Hospital for diagnosis and treatment of primary head and neck tumors.  Patients included in this study were not receiving chemo­therapy, radiotherapy, or medication at the time of sample collection.  Blood was collected by venipuncture in sterile EDTA-containing tubes and leukocytes isolated by a modification of the method of Boyum (22).  Erythrocytes were gravity-sedimented at room temperature by the addition of pyrogen-free dextran (200 kD) to a final concentration of 1.25%.  The leukocyte-rich plasma was diluted with an equal volume of HBS with EDTA, layered onto a cushion of Lymphocyte Separation Medium (Organon Teknika Corp, Durham, NC) and centrifuged at 500 x g for 5 min at room temperature.  Mononuclear leukocytes at the interface of the discontinuous gradient were collected, diluted with HBS containing EDTA and 2.5% dextran, and centrifuged at 250 x g for 5 min at room temperature.  The plate­let-rich supernatant was removed and this washing procedure repeated twice.  The PMN pellet was resuspended in HBS with EDTA and centrifuged at 5000 x g in a microcentrifuge for 2 seconds at room temperature.  Contaminating erythrocytes were removed by brief hypotonic lysis and the leukocytes washed three times in HBS with EDTA.  Mononuclear leukocytes and PMN were finally resuspended in HBS containing divalent cations and kept on ice until use.

Serum Pretreatment

Blood samples from healthy adult donors and patients with head and neck tumors were collected by venipuncture and allowed to clot overnight at 4°C.  Serum was collected by centrifugation and stored frozen in aliquots at -80°C until use.  Purified human peripheral blood PMN or mononuclear leukocytes were pretreated by incubation with 10% human serum in HBS with EDTA at 37°C for 30 min in siliconized glass test tubes.  The cells were then centri­fuged at 600 x g for 3 min at room temperature and resuspended in HBS with divalent cations at 4°C. 

Chemotaxis Assays

     N-formyl-methionyl-leucyl-phenylalanine (FMLP; Peninsula Laboratories, Belmont, CA) in concentrations ranging from 0.1 to 100 nM was placed in the bottom wells of a 48 well chemotaxis chamber (Neuroprobe, Cabin John, MD) and covered with a 5 μm pore size polyvinylpyrrolidone-free filter (Nucleopore Corp, Pleasanton, CA) as described previously (10,19).  Briefly, control and CA serum pretreated human PMN or purified mononuclear leukocytes (2-7 x 10⁴ cells) or rabbit peritoneal PMN (2 X 10⁴ cells) were loaded into the upper wells and the chambers were incu­bated at 37°C for 2 hours.  The filters were removed, fixed in methanol, stained, rinsed, dried, and mounted on glass slides using Permount.  Assays were quantitated by counting the number of cells in 5 contiguous 40X microscope fields (23).

³H-FMLP Binding and Uptake Studies

Mononuclear leukocytes (350,000/ 100 μl) or PMN (1 X 10⁶/ 100 μl) were allowed to adhere to acid-cleaned 12 mm diameter round glass coverslips in a humidified chamber at 37°C for 10 minutes.  Control or CA serum was added to each coverslip to a final concentration of 10%, and incubation continued for 30 minutes at 37°C.  Serum was removed by washing the coverslips in HBSS and these preparations studied as described below.  Cell viability remained greater than 90% throughout these experiments.  No cell loss was detected by visual inspection or cell counts of coverslips, and buffer pH was maintained between 7.4 and 7.6.

Baseline FMLP Receptor Binding at 4°C

Adherent, serum pretreated cells on coverslips were rinsed thoroughly in cold HBSS, then incubated in 75 μl of 20 nM ³H-FMLP (58 Ci/mmole; New England Nuclear, Boston, MA) for 60 min at 4°C.  The coverslips were washed briefly in 2 changes of fresh HBSS, immersed in scintillation cocktail (Biofluor; New England Nuclear, Boston, MA), and cell-associated radioactivity determined by scintillation counting (Tm Analytic Inc., Elk Grove Village, IL). 

Formylpeptide Receptor Down-Regulation

To establish a concentration curve for FMLP-induced receptor down- regulation, coverslip-adherent rabbit PMN were pretreated with either CT or CA serum, incubated with varying concentrations of unlabeled FMLP (0.1 - 20 nM) for 20 min at 37°C, rinsed thoroughly in fresh cold buffer, and then exposed to 20 nM ³H-FMLP for 60 min at 4°C.  Baseline cell-associated ³H-FMLP levels were determined in coverslip-adherent cells that had not previ­ously been exposed to unlabeled FMLP.

Receptor down-regulation over a 20 min time course was studied for cells exposed to 5 nM unlabeled FMLP at 37°C.  Subsequently, coverslips were rinsed in cold HBSS, and cell surface formylpeptide receptor expression was assessed using ³H-FMLP as described above.

Formylpeptide Receptor Reexpression on the Cell Surface

Adherent, serum pretreated cells were incubated with 5 nM unlabeled FMLP at 37°C for 20 min to down-regulate formylpeptide receptors, rinsed well with 4 changes of fresh HBSS for 10 min on ice to permit dissociation of surface-bound FMLP, and then further incubated in HBSS for 0 to 60 min at 37°C to allow receptor reexpression on the cell surface.  After these manipulations, the coverslips were rinsed well in cold HBSS, and cell surface formylpeptide receptor expression assessed as described above.

³H-FMLP Uptake

Control and CA serum pretreated adherent cells were incubated with 75 μl of 20 nM ³H-FMLP at 37°C in a humidified chamber for times ranging from 0 to 60 minutes.  At each time point, coverslips were washed in cold HBSS, immersed in BioFluor, and cell-associated radioactivity determined by scintillation counting. 

Statistical Evaluation of Data

Samples were run in triplicate and means of these triplicate groups were compared using Student's paired or unpaired t-tests.  Probability values less than 0.05 were considered significant (24).

RESULTS

Chemotaxis is Reduced in CA Patient Leukocytes and after Pretreatment of Normal Leukocytes with CA Serum

Mononuclear leukocytes from CA patients and PMN or mononuclear leukocytes pretreated with serum showed reduced chemotaxis in response to FMLP as compared to normal leukocytes or to cells similarly pretreated with CT serum.  Since this phenomenon has been described in detail elsewhere (6,7,10,11), statistics descriptive of the samples used here will only be mentioned.  Leukocytes isolated from 6 different CA patients, 17 different CA serum samples, and 8 different CT serum samples were employed.  On the average chemotaxis was reduced in CA mononuclear leukocytes (64%, 6 trials), in CA serum pretreated rabbit peritoneal PMN (40%, 6 trials), in CA serum pretreated human PMN (24%, 24 trials), and in CA serum pretreated human mononuclear leukocytes (60%, 34 trials).  Relative to that seen with CT serum, CA serum alone exhibited no significant chemotactic activity for either PMN or mononuclear leukocytes. 

Initial ³H-Formylpeptide Binding on the Cell Surface at 4°C

Cancer serum pretreated rabbit PMN bound an average of 10% less ³H-FMLP than cells similarly treated with CT serum (p<0.01; paired t-test; Figure 1A).  When human PMN and mononuclear leukocytes from normal controls were studied, no differences in ³H-FMLP binding in CT as compared to CA serum pretreated cells were observed.  In contrast, mononuclear leukocytes isolated from head and neck CA patients bound 42% less ³H-FMLP (p<0.05; Figure 1B), whereas CA patient PMN showed no significant differences in ³H-FMLP binding.

    The specificity of ³H-FMLP binding was tested by exposing coverslip-adherent PMN or mononuclear leukocytes to 20 nM ³H-FMLP in the presence of excess (20 μM) unlabeled FMLP.  Non-specific binding ranged from 1.6% to 5.3% of total ³H-FMLP binding in the experiments reported here and has been subtracted from the total binding to give receptor-specific binding.  Calculations of formylpeptide receptor numbers indicated approximately 43,000 as compared to 36,000 receptors per cell for rabbit PMN pretreated with CT as compared with CA serum.  Similar numbers of formylpeptide receptors on rabbit PMN have been reported elsewhere (25,26).

Formylpeptide Receptor Down-Regulation is not Altered by CA Serum

     1.  Formylpeptide Concentration Curve

The response of CT or CA serum pretreated rabbit PMN to challenge with unlabeled formylpeptide at 37°C was assessed over a range of FMLP concentra­tions (Figure 2).  In the absence of FMLP (i.e., 0 nM in Figure 2), there was no significant difference in ³H-FMLP binding between CT and CA serum pretreat­ed PMN.  Note that the preparation procedure for cells at this data point differed from that described for initial formylpeptide binding (previous section).  After serum treatment, an additional 20 min buffer incubation at 37°C and 10 min buffer incubation at 4°C was employed.  This extended incuba­tion time after serum treatment may have affected the FMLP binding on the cell surface such that a significant difference (as noted in previous section) no longer existed between the CT and CA serum pretreated groups.  The amount of ³H-FMLP binding varied inverse­ly with the concentration of chemoattractant used during preincubation.  With 5 nM unlabeled FMLP, ³H-FMLP counts were 27% and 25% of baseline for CT and CA sera pretreated cells, respectively.  Treatment with higher concentrations of unlabeled chemoattractant did not reduce ³H-FMLP binding beyond the level seen with 5 nM FMLP.  At each concentration of unlabeled FMLP used, ³H-FMLP binding was similar in the CT and CA serum pretreated groups.  Since a substantial reduction of cell-associated ³H-FMLP was obtained with 5 nM unlabeled FMLP, further down-regulation studies were performed using this concentration.

2.      Down-Regulation Time Course

Figure 3 shows that CA serum pretreated rabbit PMN not exposed to unla­beled FMLP (i.e., 0 time) showed 14% less ³H-FMLP binding than the correspond­ing CT serum pretreated group (p<0.02).  With increasing incubation time in unlabeled chemoattractant, cell surface ³H-FMLP binding decreased.  Clearance of receptors was rapid, with more than 50% of the original receptor-mediated binding lost after 2 min of exposure to unlabeled FMLP.  The majority of receptors (78% in CT serum pretreated cells and 75% in CA serum pretreated cells) were cleared within 5 min after initial exposure to unlabeled FMLP.  After 20 min of incubation in unlabeled FMLP, only 9% of original cell surface binding remained in both groups.  No additional significant differences were noted between CT and CA serum pretreated groups during the 20 min down-regula­tion time course. 

Formylpeptide Receptor Reexpression is not Altered by CA Serum

Coverslip-adherent rabbit PMN pretreated with CT or CA serum were first exposed to 5 nM unlabeled FMLP for 20 min, rinsed, and then further incubated in buffer for up to 60 min at 37°C to allow receptor reexpression on the cell surface.  Cells exposed to buffer alone at 37°C instead of 5 nM unlabeled FMLP showed a uniform, high level of ³H-FMLP binding throughout the 60 min observa­tion period (i.e., 100% of receptors present on cell surface).  As seen in Figure 4, PMN exposed to 5 nM unlabeled FMLP for 20 min at 37°C showed greatly decreased cell surface FMLP binding (30% and 32% of initial binding for CT or CA serum pretreated cells, respec­tively).  Upon further incubation in fresh buffer at 37°C, increasing amounts of ³H-FMLP binding were observed.  However, no significant differences in the rate or extent of formylpeptide receptor reexpression were seen in CT as compared to CA serum pretreated PMN.

Uptake of ³H-Formylpeptide at 37°C is Diminished in CA Leukocytes

When leukocytes were incubated with ³H-FMLP at 37°C, the amount of cell-associated peptide increased with time.  For rabbit PMN, a rapid increase in ³H-FMLP accumulation was evident within 2 minutes (Figure 5), but uptake by CT or CA serum pretreated cells did not differ significantly at this time point.  Thereafter, the amount of cell-associated ³H-FMLP gradually increased reaching a peak at 20 minutes, after which time it remained constant.  Cancer serum pretreated cells showed significantly reduced uptake of ³H-FMLP at 10, 20, and 40 min compared to CT serum pre­treated cells (p<0.015).

Serum pretreated human PMN and mononuclear leukocytes exhibited a slower accumu­lation of ³H-FMLP than did rabbit PMN.  During the initial 10 min of exposure to 20 nM ³H-FMLP at 37°C, uptake by CA serum pretreated human leukocytes was not significantly different from that of CT serum pretreated human leukocytes (Figure 6).  Upon further incubation, ³H-FMLP continued to accumulate such that the uptake of ³H-FMLP in the CA serum pretreated leukocytes was 20-45% less than that of the CT serum pretreated leukocytes (10 min, p<0.05; 40 min, p<0.01; 60 min, p<0.04).  Neutrophils isolated from tumor patients and CT subjects (Figure 7) exhibited patterns of ³H-FMLP uptake similar to those observed in serum pretreated leukocytes.  In these samples, uptake of ³H-FMLP was significantly reduced in CA patient PMN at 20, 40 and 60 min (20 min, p<0.001; 40 min, p<0.001; 60 min, p<0.002) as compared to that in CT patient PMN.

DISCUSSION

In the present study, CA patient mononuclear leukocytes as well as rabbit peritoneal PMN, normal human mononuclear leukocytes, and normal human PMN pretreated with 10% CA serum showed consistent reductions in formylpeptide-mediated chemotaxis when compared to leukocytes pretreated with CT serum.  Significant reductions in formylpeptide binding were observed in tumor patient mononuclear leukocytes as well as in CA serum pretreated rabbit peritoneal PMN.  However, formylpeptide receptor modulation, i.e., down-regulation from and receptor reexpression onto the cell surface, in CA serum pretreated leuko­cytes was similar to that seen in CT serum pretreated cells.  Nonetheless, significant reductions in ³H-FMLP uptake at 37°C were seen in CA serum pre­treated rabbit PMN, human PMN, human mononuclear leukocytes as compared to CT serum pre­treated leukocytes and also in tumor patient PMN as compared to PMN from normal subjects. 

Human CA serum inhibits chemotaxis of normal human PMN and monocytes (7) as well as normal guinea pig PMN (27).  Monocytes isolated from CA patient blood display a similar inhibition of chemotaxis, but chemotaxis of PMN from CA patients is not impaired (1,23,28).  The reason for this disparity is not apparent but similar findings were obtained here.  In addition, initial formylpeptide binding was reduced for rabbit PMN pretreated with CA serum as compared to CT serum pretreated cells but no significant differences were noted between CT and CA serum pretreated human leukocytes.  For leukocytes isolated directly from CA patient blood samples, formylpeptide binding was not significantly different for PMN but was significantly reduced (40%) for mononuclear leukocytes as compared to that of leukocytes from normal control subjects.  Similarly, Oostendorp et al. (18) reported decreased formylpep­tide binding by human monocytes and PMN in the presence of the p15E-related peptide, LDLLFL.  If initial formylpeptide binding is reduced, as it is for CA mononuclear leukocytes, then decreased chemotactic responsiveness may be the direct result.  Although CA serum pre­treatment resulted in decreased chemotaxis in human PMN and mononuclear leukocytes, formylpeptide binding was not altered.

To further evaluate the role of formylpeptide receptors in the cancer-associated chemotactic defect, cell surface receptor down-regulation and reexpression were examined.  The results of these studies were consistent with previous reports in which rabbit (29,30) or human (31,32) PMN were used to show that receptor-ligand complexes are removed from the cell surface by internalization result­ing in a net decrease in cell surface formylpeptide receptor expression (32‑35).  There were no significant differences between CT and CA serum pretreated PMN with regard to their formylpeptide receptor complement at any time (except at 0 time) or at any FMLP concentration employed here.  Thus, it seems that the decrease in chemotactic responsiveness in the CA serum pretreated cells is not due to alterations in the receptor down-regulation response or altered clearance of occupied cell surface formylpeptide receptors as seen under these conditions.

     Control and CA serum pretreated rabbit PMN were initially exposed to unlabeled FMLP for 20 min to induce receptor down-regulation, rinsed in fresh buffer, further incubated at 37°C for varying times to permit receptor reex­pression on the cell surface, and the binding capacity of the cells determined using ³H-FMLP.  With increasing incubation time at 37°C, a gradual increase in ³H-FMLP binding occurred but complete receptor reexpression was not observed.  Several possible technical reasons for this incomplete recovery were explored including possible cell loss, shifts in buffer pH, and nutritional deficien­cies of the incubation media, but none of these variables could account for the finding that formylpeptide receptor reexpression never exceeded 85% of the initial cell surface binding.  Nonetheless, no differences were seen in formylpeptide receptor reexpression in CT and CA serum pretreated cells.  Thus, the chemotactic defect in the CA serum pretreated group does not seem to result from altered or insufficient formylpeptide receptor reexpression.

A significant decrease in ³H-FMLP uptake was seen in rabbit PMN and in normal human PMN and mononuclear leukocytes when they were pretreated with CA serum.  A similar decrease was observed in PMN from CA patients.  This de­crease was not seen at early time points, but became apparent after 20 minutes of exposure to labeled FMLP.  Others have shown that Fc receptor-mediated phagocytosis of radiolabeled immune complexes was suppressed in human PMN and monocytes treated with fractions prepared from CA patient serum (7) and that phagocytosis was reduced in tumor patient monocytes (36).  Further, Naik et al. (37) found that fluid pinocytosis in PMN from patients with chronic myeloid leukemia was significantly reduced as compared to that of normal PMN.  Although phagocytosis, receptor-mediated endocytosis, and fluid-phase pinocyto­sis are distinctly different processes (38), it seems that CA serum may have an inhibitory effect on all three forms of endocytosis.  Accumulation of soluble ³H-FMLP at 37°C involves two concurrent cellular processes, receptor-mediated endocytosis (saturable uptake) and fluid-phase pinocytosis (non-saturable uptake).  The method employed here did not permit the contributions of each of these processes to be distinguished, but previous studies employing identical conditions have concluded that about 80% of formylpeptide uptake occurs by a receptor-mediated mechanism (29,39).  Moreover, it is evident from Figure 5 that uptake by rabbit PMN was saturable and thus primarily receptor-mediated.  Finally, preliminary experiments using ¹⁴C-polyethylene glycol, a marker for fluid phase pinocytosis, have shown no differences in uptake between CT and CA serum pretreated leukocytes after formylpeptide stimulation (unpublished data).  There is, however, an appar­ent contradiction between this finding (inhibition of receptor-mediated uptake) and the results of experiments on formylpeptide receptor down-regulation and reexpression which were not affected by CA serum pretreatment.  Although the latter are not altered in CA leukocytes under the conditions described, it is possi­ble that the effects of the CA serum may have dissipated during the lengthy incubations (60-180 min after completion of serum pretreatment) required to perform these experiments.  Inhibitory effects of serum may remain evident, however, in the former experiments since they required much shorter incubations (5-60 min) after serum treatment.  Further studies will be required to evaluate this possibility.

In general, formylpeptide receptor down-regulation and receptor reex­pression did not appear to be affected by pretreatment with cancer serum.  However, significant reductions in formylpeptide binding were observed in tumor patient mononuclear leukocytes and in cancer serum pretreated rabbit peritoneal PMN.  Reductions in chemoattractant uptake at 37°C were seen in cancer serum pre­treated rabbit PMN, human PMN, human mononuclear leukocytes and in tumor patient PMN.  These alterations in surface binding and internalization of formylpeptide may contribute to the depression of chemotaxis exhibited by these cells.  Further characterization of CA leukocytes and the effects of CA serum on leukocyte chemotaxis and fluid phase pinocytosis are in progress.

REFERENCES

 1.        Rubin RH, Cosimi AB, Goetzl EJ: Defective human mononuclear leukocyte chemotaxis as an index of host resistance to malignant melanoma. Clin Immunol Immunopathol 1976, 6:376‑388. 

  2.       Snyderman R, Pike MC, Blaylock BL, Weinstein P: Effects of neoplasms on inflammation: depression of macrophage accumulation after tumor implantation. J Immunol 1976, 116:585‑589. 

  3.       Snyderman R, Pike MC: An inhibitor of macrophage chemotaxis produced by neoplasms. Science 1976, 192:370‑372. 

  4.       Pike MC, Snyderman R: Depression of macrophage function by a factor produced by neoplasms a mechanism for abrogation of immune surveillance. J Immunol 1976, 117:1243‑1249. 

  5.       Snyderman R, Cianciolo GJ: Further studies of a macrophage chemotaxis inhibitor (MCI) produced by neoplasms: murine tumors free of lactic dehydrogenase virus produce MCI. J Reticuloendothel Soc 1979, 26:453‑458. 

  6.       Balm FJM, von Blomberg‑van deFlier BME, Drexhage HA, de Haan‑Meulman M, Snow GB: Mononuclear phagocyte function in head and neck cancer: depression of murine macrophage accumulation by low molecular weight factors derived from head and neck carcinomas. Laryngoscope 1984, 94:223‑227. 

  7.       Maderazo EG, Anton TF, Ward PA: Serum‑associated inhibition of leukotaxis in humans with cancer. Clin Immunol Immunopathol 1978, 9:166‑176. 

  8.       Cianciolo GJ, Snyderman R: Characterization of an inhibitor of monocyte function in effusions of cancer patients. Lymphokines and Thymic Hormones: Their Potential Utilization in Cancer Therapeutics. Edited by Goldstein AL, Chirigos MA. New York, Raven Press, 1981, pp. 205‑213.

  9.       Ji‑Ming W, Cianciolo GJ, Snyderman R, Mantovani A: Coexistence of a chemotactic factor and a retroviral P15E‑like related chemotaxis inhibitor in human tumor cell culture supernatants. J Immunol 1986, 137:2726‑2732. 

 10.      Walter RJ, Danielson JA, Van Alten PJ, Powell WJ: Defects in monocyte chemotaxis in patients with neoplastic disease. J Surg Res 1986, 41:215‑224. 

 11.      Cianciolo GJ: Anti‑inflammatory proteins associated with human and murine neoplasms. Biochim Biophys Acta 1986, 865:69‑82. 

 12.      Snyderman R, Seigler HF, Meadows L: Abnormalities in monocyte chemotaxis in patients with melanoma: Effects of immunotherapy and tumor removal. J Natl Cancer Inst 1977, 58:37‑41. 

 13.      Snyderman R, Meadows L, Holder W, Wells S: Abnormal monocyte chemotaxis in patients with breast cancer: Evidence for a tumor mediated effect. J Natl Cancer Inst 1978, 60:737‑740. 

 14.      Cianciolo GJ, Matthews TJ, Bolognesi DP, Snyderman R: Macrophage accumulation in mice is inhibited by low molecular weight products from murine leukemia viruses. J Immunol 1980, 124:2900‑2905. 

 15.      Cianciolo GJ, Hunter J, Silva J, Haskill TS, Snyderman R: Inhibitors of monocyte response to chemotaxins are present in human cancerous effusions and react with monoclonal antibodies to the P15(E) structural protein of retroviruses. J Clin Invest 1981, 68:831‑844. 

 16.      Tan IB, Drexhage HA, Scheper RJ, et al: Defective monocyte chemotaxis in patients with head and neck cancer. Arch Otolaryngol HN Surg 1986, 112:541‑544. 

 17.      Tan IB, Balm AJM, Snow GB, Drexhage HA: Immunosuppressive retroviral‑related factors in sera of patients with head and neck cancer. Eur Arch Otorhinolaryngol 1990, 247:387‑390. 

 18.      Oostendorp RAJ, Knol EF, Verhoeven AJ, Scheper RJ: An immunosuppressive retrovirus‑derived hexapeptide interferes with intracellular signaling in monocytes and granulocytes through N‑formylpeptide receptors. J Immunol 1992, 149:1010‑1015. 

 19.      Walter RJ, Danielson JR: Characterization of formylpeptide chemoattractant binding on neutrophils and monocytes from patients with head and neck cancer. J Natl Cancer Inst 1987, 78:61‑69. 

 20.      Harrell RA, Cianciolo GJ, Copeland TD, Oroszlan S, Snyderman R: Suppression of the respiratory burst of human monocytes by a synthetic peptide homologous to envelope proteins of human and animal retroviruses. J Immunol 1986, 136:3517‑3520. 

 21.      Gottlieb RA, Kleinerman ES, O'Brian CA, Tsujimoto S, Cianciolo GJ, Lennarz WJ: Inhibition of protein kinase C by a peptide conjugate homologous to a domain of the retroviral protein p15E. J Immunol 1990, 145:2566‑2570. 

 22.      Boyum A: Isolation of mononuclear cells and granulocytes from human blood. Scand J Clin Lab Invest 1968, 21:77‑89. 

 23.      Walter RJ, Danielson JR: Defective monocyte chemotaxis in patients with epidermoid tumors of the head and neck. Arch Otolar 1985, 111:538‑540. 

 24.      Sokal RR, Rohlf FJ: Biometry. New York, W.H. Freeman and Co., 1981, pp. 62‑97.

 25.      Mackin WM, Huang C‑K, Becker EL: The formylpeptide chemotactic receptor on rabbit peritoneal neutrophils. J Immunol 1982, 129:1608‑1611. 

 26.      Aswanikumar S, Corcoran B, Schiffman E, et al: Demonstration of a receptor on rabbit neutrophils for chemotactic peptides. Biochem Biophys Res Commun 1977, 74:810‑817. 

 27.      Miyahara T, Torisu M: Serum inhibitory factor for guinea pig macrophage and neutrophil chemotaxis found in cancer patients. Gann 1981,72:854‑861.

28.       Snyderman R, Pike MC: Defective macrophage migration produced by neoplasms: identification of an inhibitor of macrophage chemotaxis. The Macrophage in Neoplasia. Edited by Fink MA. New York, Academic Press, 1976, pp. 49‑65.

 29.      Zigmond SH, Sullivan SJ, Lauffenburger DA: Kinetic analysis of chemotactic peptide receptor modulation. J Cell Biol 1982, 92:34‑43. 

 30.      Janeczek AH, Marasco WA, Van Alten PJ, Walter RJ: Autoradiographic analysis of formylpeptide chemoattractant binding, uptake and intracellular processing by neutrophils. J Cell Sci 1989, 94:155‑168. 

 31.      Sklar LA, Finney DA, Oades ZG, Jesaitis AJ, Painter RG, Cochrane CG: The dynamics of ligand‑receptor interactions. Real‑time analysis of dissociation, and internalization of an N‑formylpeptide and its receptors on the human neutrophil. J Biol Chem 1984, 259:5661‑5669. 

 32.      Niedel JE, Kahane I, Cuatrecasas P: Receptor‑mediated internalization of fluorescent chemotactic peptide by human neutrophils. Science 1979, 205:1412‑1414. 

 33.      Sullivan SJ, Zigmond SH: Chemotactic peptide receptor modulation in polymorphonuclear leukocytes. J Cell Biol 1980, 85:703‑711. 

 34.      Jesaitis AJ, Naemura JR, Painter RG, Schmitt M, Sklar LA, Cochrane CG: The fate of the N‑formyl‑chemotactic peptide receptor in stimulated human granulocytes: Subcellular fractionation studies. J Cell Biochem 1982, 20:177‑191. 

 35.      Sklar LA, Jesaitis AJ, Painter RG, Cochrane CG: Ligand/receptor internalization: a spectroscopic analysis and a comparison of ligand binding, cellular response, and internalization by human neutrophils. J Cell Biochem 1982, 20:193‑202.

 36.      Kunz BME, Kunz RM, Albert ED: Phagocytosis of monocytes in cancer patients. Zeit Krebsforsch 1978, 91:11‑17. 

 37.      Naik NR, Advani SH, Bhisey AN: Fluid pinocytosis and esterase‑oxidase in chronic myeloid leukemic granulocytes are differentially stimulated by chemotactic peptide. Leuk Res 1992, 16:395‑401. 

 38.      Silverstein SC, Steinman RM, Cohn ZA: Endocytosis. Ann Rev Biochem 1977, 46:669‑722. 

 39.      Daukas G, Zigmond SH: Inhibition of receptor‑mediated but not fluid‑phase endocytosis in polymorphonuclear leukocytes. J Cell Biol 1985, 101:1673‑1679. 

FIGURES

Figure 1

Binding of ³H-FMLP to adherent, unstimulated leukocytes exposed to ³H-FMLP for 60 min at 4°C.

(A)  Serum pretreated normal leukocytes.  Adherent rabbit peritoneal PMN, human peripheral blood PMN, and human peripheral blood mononuclear leukocytes were pre­treated with CT or CA serum prior to exposure to ³H-FMLP.  (n=15, rabbit PMN; n=20, human PMN; n=12, human mononuclear leukocytes; mean + SD)  *  p<0.01  (B)  Patient peripheral blood leukocytes.  PMN and mononuclear leukocytes isolated from venous blood samples obtained from tumor patients or control subjects were exposed to ³H-FMLP at 4°C.  (n=3; mean + SD)  *  p<0.05 

Figure 2

Down-regulation of formylpeptide receptors in the presence of varying concen­trations of unlabeled FMLP by adherent rabbit peritoneal PMN pretreated with CT or CA serum.  Shown is a representative experiment (n=4; mean + SD). 

Figure 3

Time course of formylpeptide receptor down-regulation for adherent rabbit peritoneal PMN pretreated with CT or CA serum and incubated with 10 nM unla­beled FMLP.  Shown is a representative experiment (n=4; mean + SD).

Figure 4

 

Formylpeptide receptor recovery after down-regulation for adherent rabbit peritoneal PMN pretreated with CT or CA serum, exposed to 5 nM unlabeled FMLP, rinsed, and further incubated in buffer at 37°C.  Shown is the percent of cell surface receptors expressed as compared to cell samples not exposed to unla­beled FMLP (i.e., 100% receptor expression)  (n=3; mean + SD).

Figure 5

Uptake of ³H-FMLP (20 nM) at 37°C by adherent rabbit PMN pretreated with CT or CA serum.  Uptake of ³H-FMLP in the CA serum pretreated group is significantly reduced at 10 min (p<0.003), 20 min (p<0.001) and 40 min (p<0.015).  Shown is a representative experiment (n=8).

Figure 6

Uptake of ³H-FMLP at 37°C in the presence of 20 nM ³H-FMLP by adherent normal human leukocytes pretreated with CT or CA serum.  (A)  Uptake in the CA serum pretreated cells is significantly reduced at 40 min (p<0.01) and 60 min (p<0.04).  Shown is a representative experiment (n=6; mean + SD).  (B)  Uptake in CA serum pretreated mononuclear leukocytes is significantly reduced at 40 min (p<0.01) in the CA serum pretreated cells.  Shown is a representative experiment (n=5; mean + SD).  

Figure 7

Uptake of ³H-FMLP at 37°C in the presence of 20 mM ³H-FMLP by adherent PMN isolated from control subjects and tumor patients.  Uptake in the CA patient PMN is significantly reduced at 20 min (p<0.001), 40 min (p<0.001) and 60 min (p<0.002).  Shown is a representative experiment (n=3; mean + SD).

FYI.  The content on this blog is copyright protected by the author.  Feel free to read, copy, and disseminate the studies described here.  When it is appropriate to recognize and attribute the work of others, please indicate the origin of the work by citing this blog URL as an electronic web citation .

Growth of Human Keratinocytes on ADM in Vitro

May 1, 2012

This study was an attempt to use the acellular dermal matrix material that we had developed earlier into a substitute for full-thickness skin.  ADM could be used as a dermal filler or substitute in situations where the dermis was damaged or absent but, by itself, ADM lacked an epithelial barrier to infection and fluid loss.  Here we used ADM as the substrate for growing human keratinocytes in culture. A patent layer of keratinocytes would provide an effective epithelial barrier over the ADM.

However, keratinoctes are fairly difficult to grow due to their stringent growth requirements.  They can be easily and irreversibly damaged by the tryptic enzymes used for cell transfers, require carefully controlled soluble calcium levels for optimal growth and differentiation, must be handled carefully during transfers to minimize damage, require a number of growth factors and hormones to achieve reasonable or good survival and growth. 

This sensitivity to growth conditions complicated this study considerably.  We added a range of different attachment factors, growth factors, or extracellular matrix components to the ADM in an effort to promote keratinocytes attachment and growth on the ADM substrate.  The following paper describes the results of these experiments.

The full paper and a link to the PDF are shown below.

Growth of Human Keratinocytes on ADM in Vitro

Growth of Human Keratinocytes on 

Acellular Dermal Matrix in Vitro

By

Lawrence J. Jennings, MD¹, Evangeline Z. DeSagun, MD²,

Marella Hanumadass, MD¹, and Robert J. Walter, PhD³ 

¹ Burn Center, Cook County Hospital, Chicago, IL 60612

² Department of Pathology, Cook County Hospital, Chicago, IL 60612

 ³ Department of Surgery, Cook County Hospital and Department of General Surgery,          Rush University Medical Center, Chicago, IL 60612

Running Title:    Keratinocyte growth on ADM

Keywords:  keratinocytes, acellular dermal matrix, matrigel, dermis, collagen, laminin, chondroitin sulfate, dermatan sulfate, hyaluronate, fibronectin, heparin, polylysine, fetal calf serum, artificial skin, wound healing, burn

Correspondence should be addressed to:

Robert J. Walter, PhD

Division of Surgical Research

Department of Surgery

Cook County Hospital

627 South Wood Street

Chicago, Illinois   60612 

Telephone (312) 633-7237;  FAX (312) 633-8347

E-mail: rwalter@rush.edu

ABSTRACT

            A composite skin substitute composed of acellular dermal matrix (ADM) covered by a layer of human keratinocytes (KCs) would be very useful in treating burns or other injuries.  To determine the factors necessary to support KC attachment and growth on human ADM, we treated ADMs with different extracellular matrix components and cell attachment factors, plated KCs onto the treated ADM, and then cultured the KCs for three or ten days.  At these times, ADMs were frozen, cryosectioned, and assessed for KC attachment and growth.  ADM prepared in the presence of azide was mildly toxic to KCs that came into direct contact with it even after it had been extensively washed.  ADM prepared with antibiotics provided a much better substrate for KC attachment and growth, but growth was still not optimal.  Despite pretreating the ADM with a range of basement membrane constituents (laminin, Matrigel, collagen IV), extracellular matrix components (collagen I, chondroitin or dermatan sulfate, hyaluronate, fibronectin), or other attachment factors (heparin, polylysine), keratinocyte growth was not significantly improved.  Pretreatment of the ADM with fetal calf serum (20% or 100%) or co-culture with 2%fetal calf serum, fibroblast conditioned medium, or 3T3 cells significantly promoted keratinocyte attachment to and growth on ADM such that confluent monolayers formed.

INTRODUCTION

In extensively burned patients, prompt replacement of damaged skin is essential to limit the morbidity and mortality associated with these injuries.  Since donor sites for autografts are limited and the donor sites themselves become a source of long-term morbidity, alternative sources of non-immunogenic skin and functional skin substitutes are being developed (1-6).  In this pursuit, cultured epithelial autografts derived from the growth of autogenic keratinocytes (KCs) in vitro (7,8) and several types of dermal substitutes have been used with varying degrees of success (9-14).  Dermal substitutes composed of gelled type I collagen (6,15), type I collagen mixed with glycosaminoglycan (16), or fibrin (17) is conducive to the attachment of human KCs, their subsequent proliferation, and the formation of differentiated multilayers.  In addition, many basement membrane and extracellular matrix components (ECM) including type IV collagen, laminin, fibronectin, and RGD tripeptide (arg-gly-asp) (18,19) have been found to improve human KC attachment and sometimes their proliferation on a variety of substrates used for tissue culture.

Using de-epidermized dermis, Krejci et al. (20) found that KCs from foreskin explants readily grew out onto the papillary, but not the reticular, dermal surface in vitro.  This suggested that one or more components of the basement membrane might be required for KC attachment and growth on native dermis.  However, pretreatment of the dermis with fibronectin or type IV collagen did not improve KC outgrowth onto the reticular surface of de-epidermized dermis.  Ralston et al. (21) have also shown the importance of basement membrane antigens in facilitating KC attachment and growth on de-epidermized dermis.  While de-epidermized dermis is a useful substrate for studying KC growth and differentiation, it seems to retain a variety of cell-associated antigens that may make it unsuitable for implantation into an immunocompetent host (22).  A more thoroughly decellularized material, acellular dermal matrix (ADM), is derived from human cadaver skin that has been processed to remove the epidermis and all cellular components leaving only the connective tissue dermal matrix.  It has been found to be very weakly- or non-antigenic and is an effective dermal substitute when used in conjunction with onlay grafts of ultrathin split-thickness skin (3,23,24).  However, ADM is a poor substrate for KC growth in vitro.

Rennekampff et al. (25) found that human KCs attached poorly to AlloDerm®, one type of ADM, in vitro and that added fibronectin did not improve this poor adherence.  Further, we have found that ADMs prepared by treating skin using either the NaCl/ sodium dodecyl sulfate method (like AlloDerm®) or the dispase/Triton X-100 method (3) serve as poor substrates for KC growth in serum-free, low-Ca⁺⁺ medium (26).  We have also reported that several ECMs including glycosaminoglycans (chondroitin, dermatan, and keratan sulfates, and hyaluronic acid), components of the basal lamina (collagen types IV and VII, laminin), and cell attachment factors (e.g., fibronectin) are partially depleted or absent from these ADMs (14).  Because many of these tissue components are known to promote epithelial attachment to underlying connective tissue and to support epithelial cell proliferation, we hypothesized that KC attachment and growth on ADM might be improved by replenishing one or more of these factors or by employing other agents to promote cell attachment.  Here we report the effects of a variety of crude and purified ECMs, growth factors, and cell attachment factors on KC attachment to and proliferation on dispase/ Triton human ADM prepared in the presence of either azide or antibiotics. 

ABBREVIATIONS USED

acellular dermal matrix, ADM; extracellular matrix components, ECMs; fetal calf serum, FCS; keratinocyte, KC; keratinocyte growth medium, KGM; phosphate buffered saline, PBS

MATERIALS AND METHODS

Preparation of ADM

Normal human skin, 0.012 inches thick, obtained from cadavers using a dermatome (Padgett Electro-Dermatome, Padgett Instruments, Inc., Kansas City, MO) was washed in RPMI-1640 containing 10% human serum, and then frozen at -80°C.  Tissue was obtained according the ethical guidelines established in the 1975 Declaration of Helsinki and the Institutional Review Board.  Cryopreserved skin was thawed rapidly at 37°C in saline and was then treated with 2.5 units/ml dispase II (Boehringer Mannheim, Indianapolis, IN) in phosphate buffered saline (PBS) containing 0.2 mM CaCl₂ at 4°C for 24 hours to remove the epidermis and other cellular components from the dermal matrix.  Subsequently, the dermal matrix was incubated in buffered 0.5% Triton X-100 (U. S. Biochemical Corp., Cleveland, OH) for 24 hours at room temperature with continuous shaking.  Dispase-Triton ADM was then extensively washed with PBS and stored at 4°C until use.  All solutions were filter-sterilized and all procedures were performed aseptically.  Either sodium azide (0.02% w/v) or a cocktail of antibiotics (300 U/ml penicillin, 0.3 mg/ml streptomycin, 0.75 μg/ml fungizone, 50 mg/ml gentamycin) was present at all times during both of these steps to deter microbial growth (3). 

ADM Treatment and Keratinocyte Culture

ADMs were washed extensively with sterile saline, cut into 10X10 mm pieces, treated with ECMs or attachment factors for 24 h at room temperature, rinsed thoroughly with PBS, attached to sterile stainless steel mesh, and placed into 24-well culture plates.  The following ECMs and attachment factors were used: laminin (12 μg/cm²), fibronectin (10 μg/cm²), Matrigel (60 μl/cm²), fibrin (0.5 mg/cm²), collagen type I (0.25 mg/cm²), collagen type IV (2.5 μg/cm²), heparin (5 mg/cm²), chondroitin sulfate (5 mg/cm²), dermatan sulfate (0.5 mg/cm²), hyaluronic acid (0.2 mg/cm²), fetal calf serum (FCS; 2%, 20%, 100%), and polylysine (0.1%, 0.001%).  Types I and IV collagen, Matrigel, and laminin were from Collaborative Biomedical, Bedford, MA; hyaluronic acid from Seikagaku, Tokyo, Japan; plasma fibronectin, polylysine, fibrinogen, thrombin, chondroitin sulfate, and dermatan sulfate from Sigma Chemicals, St. Louis, MO; heparin from US Biochemicals, Cleveland, OH; and FCS from Gibco BRL, Grand Island, NY.

Human KCs obtained from foreskin were maintained in KGM supplemented with bovine pituitary extract (Clonetics, Temecula, CA) and were trypsinized and plated into 24-well plates (100,000 cells/ well) containing treated ADMs.  Cultures were continued for 3 or 10 days in KGM alone, in KGM supplemented with 10% human fibroblast-conditioned medium or 2% FCS or in KGM with co-cultured 3T3 fibroblasts.

Tissue Preparation and Staining

ADMs were removed from culture vessels, embedded in Tissue Freezing Medium (Triangle Biomedical, Durham, NC), frozen, and cut in cross-section at a thickness of 8 mm using an IEC Minotome cryostat.  Sections were picked up on chrome-albumin subbed glass slides and allowed to dry at room temperature.  Dried cryosections were fixed in formalin and 95% ethanol (1:9), rinsed in water, and stained using H&E.  The nuclei of KCs attached to the ADM were counted in twenty 40X microscope fields (the field diameter was 0.45 mm) and this number was then averaged and converted to cells per linear mm of ADM cross-section (cells/ mm).

Statistical Analysis

Cell counts for each group were averaged and means ± SEM calculated.  Unpaired t-tests or ANOVA were used to compare groups and p values less than 0.05 were considered significant.

RESULTS

As seen in Figure 1, all of the ADMs prepared in the presence of azide showed relatively few attached KCs.  KC attachment (Figures 2 and 3) for the ADMs pretreated with 100% FCS, fibroblast-conditioned medium, or grown in the presence of 2% FCS showed improved KC attachment and growth after 3 days in culture (8.6, 10.7, and 10.5 cells per mm of cross-sectioned ADM, respectively) as compared to untreated ADM (5.4 cells per mm) but these differences were not significant.  Polylysine did not affect cell attachment but reduced cell viability.  Laminin, chondroitin sulfate, dermatan sulfate, heparan sulfate, and collagen IV did not significantly affect KC attachment and growth on ADM.  All groups tested showed fewer attached KCs after 10 days in culture than after 3 days (p<0.001, ANOVA).  On the tenth day of culture, the cell numbers in the 2% FCS group were significantly (p=0.03, t-test) greater than those of the controls.

Greatly improved cell attachment and growth (Figure 4) were noted in ADM processed in the presence of antibiotics (i.e., without azide) as compared to ADM processed in the presence of azide (p<0.001; ANOVA).  Nevertheless, fewer cells were present after 10 days in culture for most groups (Figures 5 and 6; e.g., 7.8 vs. 2.0 cells/mm for the control group on days 3 and 10, respectively).  However, for the KCs grown on ADM pretreated with FCS (20% or 100%), or cultured in the presence of human fibroblast-conditioned medium or 2% FCS (Figure 6), cell proliferation continued after the third day.  As a result, continuous KC monolayers were observed on the ADM by the tenth day (e.g., 10.9 cells/mm at 3 days and 23.5 cells/mm at 10 days for the 20% FCS group).  By the tenth day in culture, the cell numbers in the 2% FCS, 20% FCS, human fibroblast conditioned medium, and co-cultured 3T3 cell groups were significantly (p<0.003, p<0.001, p<0.003, p<0.04, respectively; t-tests) greater than those of the control.  ADMs treated with other attachment factors showed fewer attached KCs after 10 days in culture than after 3 days. 

DISCUSSION

Despite extensive washing, residual azide may have remained associated with ADMs prepared in the presence of azide, resulting in poor KC attachment and survival in the 3 and 10 day groups (see Figures 2 and 3).  Interestingly, the only KCs affected were those directly in contact with the ADM.  KCs that attached to the bottom of the ADM-containing culture wells proliferated normally during the ten day culture period.  Thus, the azide or other deleterious substance(s) was not released into the culture medium and affected only cells in direct contact with the ADM.  Cell numbers declined for most groups on azide-treated ADM between the third and tenth days in culture.  To some extent, this can be attributed to the aforementioned azide effect, but a similar decline was also seen for many of the groups in which KCs were grown on azide-free ADM (e.g., untreated, fibronectin, chondroitin sulfate, dermatan sulfate, etc.)(see Figure 5).  There are several possible reasons that such a varied assortment of basement membrane components, glycosaminoglycans, and other ECMs would fail to improve KC attachment to and growth on ADM.  These supplements may not have bound to the ADM initially or they may have bound but: 1) subsequently dissociated from the ADM, 2) the ADM may have exerted an overriding inhibitory effect on KC growth, or 3) do not benefit KC growth.

To assure initial binding of the supplements, they were applied at the concentrations and under conditions similar to those described elsewhere or as recommended by the manufacturers.  However, most other studies that use ECMs to enhance cell attachment or growth in culture permit them to air-dry onto plastic tissue culture substrates or dermal matrices (see e.g., (19,27,28)) or incorporate them into semi-synthetic matrices during fabrication (16,29).  The latter method was not feasible here and air-drying was deemed undesirable because of the attendant potential for denaturation or modification of ADM components due to surface tension forces.  While such alterations might not adversely affect the results of the present in vitro studies, denatured ADMs used for implantation would likely evince greater antigenicity and less stability, thereby adversely affecting wound healing.  Instead, in the present study, we sought to exploit the known binding affinities of the ECM materials for type I collagen and other components of the hydrated, native ADM in order to effect their attachment to the dermal matrix.  To assess the initial binding to and retention by ADM, we used fluorescein-conjugated anti-fibronectin to detect fibronectin on cryosectioned ADM using fluorescence microscopy.  This confirmed that supplemental fibronectin did bind to the ADM under the conditions used here for the fibronectin, FCS (2%, 20%, 100%) co-cultured 3T3 cells, and fibroblast conditioned medium groups and that this fibronectin remained bound for the entire ten day culture period (data not shown).  Based on this finding and the known binding affinities of the various supplements used, these supplements very likely bound initially to the ADM, but some may have subsequently dissociated from the ADMs or were degraded during extended cell culture thereby abrogating any potentially beneficial effects to KC proliferation.  Indeed, Hanthamrongwit et al. (30)noted that chondroitin sulfate dissociated rapidly after it had been applied to collagen sponges used as in vitro KC substrates. 

Components of the basement membrane including laminin-5 (17), type IV collagen (17,18), and Matrigel (a mixture of type IV collagen, laminin, heparan sulfate, and entactin) (18,28) have each been shown to promote KC attachment and proliferation on tissue culture plastic and on some types of dermal matrices (21).  Related connective tissue components such as type I collagen gels, plasma fibronectin, and RGD have shown variable effects on KC attachment ranging from inhibition to stimulation (18,19,28,29).  However in the present study, as in that of Krejci et al. (20) where type IV collagen and fibronectin were tested on second-cut dermis, each of these compounds had little effect on KC attachment and growth.  Glycosaminoglycans such as chondroitin sulfate, dermatan sulfate, and hyaluronate are also known to improve in vitro KC attachment and proliferation (2,16,30-32) and fibrin glue has been found to be effective as a substrate for KC growth (17,33).  However, as with the present study, Shakespeare and Shakespeare (34) reported that fibrin appeared to inhibit KC attachment to dermal collagen.  The above-mentioned basement membrane components and ECMs may have been ineffective in the present system because most of them form gels at the concentrations employed for cell culture and these gels may infiltrate poorly into the ADM.  As a result, the gelled material may readily dissociate or detach from the ADM together with any attached KCs.  Our observations of cryostat sections tend to confirm this for type I collagen, Matrigel, hyaluronate, and chondroitin sulfate.  Further studies to quantify the extent of the initial binding to and dissociation from ADM of these supplements are needed and are in progress.

The effects of polylysine and heparin on KC growth on ADM were also evaluated.  These agents were studied because of their known ability to bind to surfaces by electrostatic interactions (polylysine) or through specific binding domains (heparin) on both cell surfaces and collagen.  KCs grown on ADMs pretreated with polylysine at high concentration (0.1%) were non-viable probably due to the continued release of residual polylysine from the ADM since other cells in the same wells, attached to the bottom of the culture plate, were also killed.  At lower concentrations, these substances were found to have no effect on KC binding and proliferation.

KCs grown on azide-free ADM cultured in the presence of human fibroblast-conditioned medium (which also contained FCS at a final concentration of 2%) developed into confluent monolayers on the ADM by the tenth day in culture.  Under similar conditions, we have also seen that multilayers developed when these monolayer cultures were moved to the air-liquid interface (unpublished data).  For the purposes of the present study, the formation of monolayers was used as the end-point.  It is not entirely surprising that fibroblast-conditioned medium and co-cultured 3T3 cells enhance KC growth on ADM since the latter is the classical method for stimulating KC growth in vitro (35) and this method has also succeeded in improving KC growth on dermal substitutes in other studies (18,20,21,28,36).  However, even pretreatment of the ADM with FCS (20% or 100%) was sufficient to permit extensive KC attachment, proliferation, and monolayer development by the tenth day in culture.  Because plasma fibronectin is one of the most abundant cell attachment factors found in serum, we tested the effect of purified fibronectin on KC attachment to and proliferation on ADM, but as described above, it had no effect.  Thus, we cannot determine precisely how FCS benefits KC attachment and proliferation in this system.  FCS may exert a non-specific blocking effect, coating the ADM with protein and masking sites or substances that would otherwise inhibit KC attachment and growth.  Alternatively, in experiments where 2% FCS was present during the entire ten day culture period, FCS may provide a continuous source of factors (e.g., plasma fibronectin, mitogens, nutritional factors) that enhance KC attachment and proliferation.  Because KC survival was reduced in all other treatment groups between the third and tenth days in culture, we speculate that either the KCs underwent apoptosis due to inadequate cell-substrate interaction or the nutritional requirements for KCs growing on ADM are altered such that KGM is not sufficient to maintain viability.

In conclusion, it seems that the presence of azide during ADM preparation is detrimental to the subsequent growth of KCs on ADM, but that KC attachment and proliferation are greatly improved on ADM prepared in the presence of antibiotics rather than azide.  Nonetheless, a variety of purified ECMs and other cell attachment factors known to be effective in promoting cell attachment or growth in other culture systems were ineffective in promoting KC growth on ADM.  Alternative methods for applying these supplements to the ADM such as chemical cross-linking, lyophilization, or continuous application throughout the culture period are currently under study.  The presence of FCS, co-cultured fibroblasts, fibroblast-conditioned medium, or pretreatment of the ADM with FCS resulted in good KC attachment and growth on ADM and resulted in growth of confluent KC monolayers on ADM.  These studies show that dispase/ Triton ADM can be a good substrate for KC growth and indicate that it has future potential for use in the production of KC-ADM composites for use in implantation.

REFERENCES

   1.   Yannas IV, Burke JF. Design of an artificial skin. 1.  Basic design principles. J Biomed Mater Res. 1980;14:65-81.

   2.   Hansbrough JF, Boyce ST, Cooper ML, Foreman T. Burn wound closure with cultured autologous keratinocytes and fibroblasts attached to a collagen-glycosaminoglycan substrate. JAMA. 1989;262:2125-30.

   3.   Takami Y, Matsuda T, Yoshitake M, Hanumadass M, Walter RJ. Dispase/detergent treated dermal matrix as a dermal substitute. Burns. 1996;22:182-90.

   4.   Livesey SA, Herndon DN, Hollyoak MA, Atkinson YH, Nag A. Transplanted acellular allograft dermal matrix.  Potential as a template for the reconstruction of viable dermis. Transplantation. 1995;60:1-9.

   5.   Boyce ST, Goretsky MJ, Greenhalgh DG, Kagan RJ, Rieman MT, Warden GD. Comparative assessment of cultured skin substitutes and native skin autograft for treatment of full-thickness burns. Annals Surg. 1995;222:743-52.

   6.   Bell E, Ehrlich HP, Buttle DJ, Nakatsuji T. Living tissue formed in vitro and accepted as skin-equivalent tissue of full thickness. Science. 1981;211:1052-54.

   7.   Gallico GG, O'Connor NE, Compton CC, Kehinde O, Green H. Permanent coverage of large burn wounds with autologous cultured human epithelium. New Engl J Med. 1984;311:448-51.

   8.   Munster AM. Cultured skin for massive burns.  A prospective, controlled trial. Annals Surg. 1996;224:372-77.

   9.   De Vries HJC, Middelkoop E, Mekkes JR, Dutrieux RP, Wildevuur CH, Westerhof W. Dermal regeneration in native non-cross-linked collagen sponges with different matrix molecules. Wound Repair Regen. 1994;2:37-47.

10.   Boyce ST, Hansbrough JF. Biologic attachment, growth and differentiation of cultured human epidermal keratinocytes on a graftable collagen and chondroitin-6-sulfate substrate. Surgery. 1988;103:421-31.

11.   McKay I, Woodward B, Wood K, Navsaria HA, Hoekstra H, Green C. Reconstruction of human skin from glycerol-preserved allodermis and cultured keratinocyte sheets. Burns. 1994;20:S19-S22.

12.   Wilkins LM, Watson SR, Prosky SJ, Meunier SF, Parenteau NL. Development of a bilayered living skin construct for clinical applications. Biotech Bioengin. 1994;43:747-56.

13.   Eaglstein WH, Iriondo M, Laszlo K. A composite skin substitute (Graftskin) for surgical wounds.  A clinical experience. Dermatol Surg. 1995;21:839-43.

14.   Walter RJ, Matsuda T, Reyes HM, Walter JM, Hanumadass M. Characterization of acellular dermal matrices (ADMs) prepared by two different methods. Burns. 1998;24:104-13.

15.   Wang HJ, Wan HL, Yang TS, Wang DS, Chen TM, Chang DM. Acceleration of skin graft healing by growth factors. Burns. 1996;22:10-14.

16.   Compton CC, Butler CE, Yannas IV, Warland G, Orgill DP. Organized skin structure is regenerated in vivo from collagen-GAG matrices seeded with autologous keratinocytes. J Invest Dermatol. 1998;110:908-16.

17.   Meana A, Iglesias J, Del Rio M, Larcher F, Madrigal B, Fresno MF et al. Large surface of cultured human epithelium obtained on a dermal matrix based on live fibroblast-containing fibrin gels. Burns. 1998;24:621-30.

18.   Dawson RA, Goberdhan NJ, Freedlander E, MacNeil S. Influence of extracellular matrix proteins on human keratinocyte attachment, proliferation and transfer to a dermal wound model. Burns. 1996;22:93-100.

19.   Rennekampff HO, Kiessig V, Griffey S, Greenleaf G, Hansbrough JF. Acellular human dermis promotes cultured keratinocyte engraftment. J Burn Care Rehabil. 1997;18:535-44.

20.   Krejci NC, Cuono CB, Langdon RC, McGuire J. In vitro reconstitution of skin:  Fibroblasts facilitate keratinocyte growth and differentiation on acellular reticular dermis. J Invest Dermatol. 1991;97:843-48.

21.   Ralston DR, Layton C, Dalley AJ, Boyce S, Freedlander E, Mac Neil S. The requirement for basement membrane antigens in the production of human epidermal/dermal composites in vitro. Brit J Dermatol. 1999;140:605-15.

22.   Prunieras M. The reconstruction of skin:  How and why. Intl J Dermatol. 1990;29:553-56.

23.   Walter RJ, Jennings LJ, Matsuda T, Reyes HM, Hanumadass M. Dispase/Triton treated acellular dermal matrix as a dermal substitute in rats. Curr Surg. 1997;54:371-74.

24.   Wainwright D, Madden M, Luterman A, Hunt J, Monafo W, Heimbach D et al. Clinical evaluation of an acellular allograft dermal matrix in full-thickness burns. J Burn Care Rehabil. 1996;17:124-36.

25.   Rennekampff HO, Hansbrough JF, Woods V, Kiessig V. Integrin and matrix molecule expression in cultured skin replacements. J Burn Care Rehabil. 1996;17:213-21.

26.   Jennings LJ, DeSagun E, Hanumadass M, Walter RJ. Growth of human keratinocytes on acellular dermal matrix in vitro. J Burn Care Rehabil. 1998;19:S171-S171.

27.   Daniels JT, Kearney JN, Ingham E. An investigation into the potential of extracellular matrix factors for attachment and proliferation of human keratinocytes on skin substitutes. Burns. 1997;23:26-31.

28.   Maruguchi T, Maruguchi Y, Suzuki S, Matsuda K, Toda KI, Isshiki N. A new skin equivalent:  Keratinocytes proliferated and differentiated on collagen sponge containing fibroblasts. Plastic Reconstruc Surg. 1994;93:537-43.

29.   Cooper ML, Hansbrough JF, Foreman TJ. In vitro effects of matrix peptides on a cultured dermal-epidermal skin substitute. J Surg Res. 1990;48:528-33.

30.   Hanthamrongwit M, Reid WH, Grant MH. Chondroitin-6-sulphate incorporated into collagen gels for the growth of human keratinocytes:  The effect of cross-linking agents and diamines. Biomaterials. 1996;17:775-80.

31.   Matsuda K, Suzuki S, Isshiki N, Yoshioka K, Okada T, Ikada Y. Influence of glycosaminoglycans on the collagen sponge component of a bilayer artificial skin. Biomaterials. 1990;11:351-55.

32.   Myers SR, Grady J, Soranzo C, Sanders R, Green C, Leigh IM et al. A hyaluronic acid membrane delivery system for cultured keratinocytes:  Clinical "take" rates in the porcine kerato-dermal model. J Burn Care Rehabil. 1997;18:214-22.

33.   Horch RE, Bannasch H, Kopp J, Andree C, Stark GB. Single-cell suspensions of cultured human keratinocytes in fibrin-glue reconstitute the epidermis. Cell Transplantation. 1998;7:309-17.

34.   Shakespeare VA, Shakespeare PG. Growth of cultured human keratinocytes on fibrous dermal collagen:  A scanning electron microscope study. Burns. 1987;13:343-48.

35.   Rheinwald JG, Green H. Serial cultivation of human epidermal keratinocytes: The formation of keratinizing colonies from single cells. Cell. 1975;6:331-40.

36.   Medalie DA, Tompkins RG, Morgan JR. Evaluation of acellular human dermis as a dermal analog in a composite skin graft. ASAIO Journal. 1996;42:M455-M462.

FIGURES

  Figure 1

Human KCs growing on the surface of human ADM after 3 days (left) or 10 days (right) in culture.  ADM was prepared in the presence of azide, pretreated with the substances shown for 24 hrs, washed, and then KCs were plated onto the ADM.  For some of the treatment groups (10% human fibroblast conditioned medium, 2% FCS, or co-cultured 3T3 cell groups), the KCs were maintained for the entire 3 or 10 day period in the presence of these additional factors.  After 3 or 10 days in culture, ADMs were cryosectioned and photographed. H&E stain.  Magnification bar = 400 μm

 

 Figure 2

Human azide-prepared ADM was pretreated for 24 hr as shown, washed in buffer, human keratinocytes were plated onto the treated ADM in 24-well plates (100,000/ well), and the cultures continued for 3 (open bars) or 10 (cross-hatched bars) days in KGM.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts ± SEM.  The number associated with each bar represents the number of experiments (n) performed using each compound; bars for 3 and 10 day groups are overlapping not stacked.

 Figure 3

Human azide-prepared ADM was pretreated for 24 hr as shown, washed in buffer, human keratinocytes were plated onto the ADM in 24-well plates (100,000/ well), and the cultures continued for 3 (open bars) or 10 (cross-hatched bars) days in KGM.  Alternatively, keratinocytes were maintained in the presence of 10% human fibroblast (hu Fb) conditioned medium (which contained 2% FCS), 2% FCS alone, or co-cultured 3T3 cells for the entire 3 or 10 day period.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts ± SEM.  The number associated with each bar represents the number of experiments (n) performed using each compound; bars for 3 and 10 day groups are overlapping not stacked.

Figure 4

Human KCs growing on the surface of human ADM after 3 days (left) or 10 days (right) in culture.  ADM was prepared in the absence of azide, pretreated with the substances shown for 24 hrs, washed, and then KCs were plated onto the ADM.  For some of the treatment groups (10% human fibroblast conditioned medium, 2% FCS, or co-cultured 3T3 cell groups), the KCs were maintained for the entire 3 or 10 day period in the presence of these additional factors.  After 3 or 10 days in culture, ADMs were cryosectioned and photographed. H&E stain.  Magnification bar = 400 μm

 

Figure 5

Human ADM prepared in antibiotics was pretreated for 24 hr as shown, washed in buffer, human keratinocytes were plated onto the ADM in 24-well plates (100,000/ well), and the cultures continued for 3 (open bars) or 10 (cross-hatched bars) days in KGM.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts ± SEM.  The number associated with each bar represents the number of experiments (n) performed using each compound; bars for 3 and 10 day groups are overlapping not stacked.

 Figure 6

Human ADM prepared in antibiotics was pretreated for 24 hr as shown, washed in buffer, human keratinocytes were plated onto the ADM in 24-well plates (100,000/ well), and the cultures continued for 3 (open bars) or 10 (cross-hatched bars) days in KGM.  Alternatively, keratinocytes were maintained in the presence of 10% human fibroblast (hu Fb) conditioned medium, 2% FCS, or co-cultured 3T3 cells for the entire 3 or 10 day period.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts ± SEM.  The number associated with each bar represents the number of experiments (n) performed using each compound; bars for 3 and 10 day groups are overlapping not stacked.

 TABLE 1

Keratinocyte Attachment and Growth on Azide-Sterilized ADM

Effect of Basement Membrane and Extracellular Matrix Components

Treatment	Human keratinocytes/ mm of ADM cross-section
ADM with azide	3 day cultures	n	10 day cultures	n
control, untreated	5.4 ± 1.7	5	2.7 ± 0.7	5
laminin	3.6 ± 0.5	3	0.8 ± 0.2	3
plasma fibronectin	2.1 ± 0.2	3	1.7 ± 0.3	3
Matrigel	1.0 ± 0.2	2	0.8 ± 0.3	2
collagen type IV	2.1 ± 0.5	2	0.5 ± 0.2	2
collagen type I	7.8 ± 3.3	3	1.4 ± 0.7	3
chondroitin sulfate	1.4 ± 0.5	2	0.4 ± 0.1	2
dermatan sulfate	4.6 ± 1.0	2	0.2 ± 0.1	2
hyaluronic acid	3.7 ± 0.8	2	1.2 ± 0.5	2

Human ADM was pretreated for 24 hr with the compounds shown above, washed in buffer, human keratinocytes were plated onto them in 24-well plates (100,000/ well), and the cultures continued for 3 or 10 days in KGM.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts from the indicated number (n) of  experiments.  Means ± SEM.

TABLE 2

Keratinocyte Attachment and Growth on Azide-Sterilized ADM

Effect of Attachment Agents

Treatment	Human keratinocytes/ mm of ADM cross-section
ADM with azide	3 day cultures	n	10 day cultures	n
control, untreated	5.4 ± 1.7	5	2.7 ± 0.7	5
FCS - 2%	10.5 ± 4.5	3	8.0 ± 1.9 (p=0.03)	3
FCS - 20%	4.4 ± 0.7	5	3.6 ± 0.5	5
FCS - 100%	8.6 ± 2.3	2	4.6 ± 1.2	2
polylysine - 0.1%	5.3 (non-viable)	4	5.0 (non-viable)	4
polylysine - 0.001%	6.8 ± 1.2	4	5.0 ± 1.3	4
heparin	0.9 ± 0.5	2	0.4 ± 0.5	2
hu Fb conditioned medium	10.7 ± 2.1	3	4.7 ± 1.6	3
co-cultured 3T3 cells	8.8 ± 2.8	5	4.8 ± 1.8	5

Human ADM was pretreated for 24 hr with the compounds shown above, washed in buffer, human keratinocytes were plated onto them in 24-well plates (100,000/ well), and the cultures continued for 3 or 10 days in KGM.  Alternatively, keratinocytes were maintained in the presence of 10% human fibroblast (hu Fb) conditioned medium, 2% FCS, or co-cultured 3T3 cells for the entire 3 or 10 day period.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts from the indicated number (n) of  experiments.  Means ± SEM.

TABLE 3

Keratinocyte Attachment and Growth on Antibiotic-Sterilized ADM

Effect of Basement Membrane and Extracellular Matrix Components

Treatment	Human keratinocytes/ mm of ADM cross-section
ADM without azide	3 day cultures	n	10 day cultures	n
control, untreated	7.8 ± 1.7	9	2.0 ± 0.7	9
laminin	6.4 ± 1.5	2	1.1 ± 0.2	2
plasma fibronectin	7.8 ± 2.7	9	2.0 ± 0.5	9
Matrigel	4.7 ± 1.6	2	3.3 ± 0.6	2
collagen type IV	6.1 ± 1.2	2	1.5 ± 0.6	2
collagen type I	6.7 ± 1.5	2	3.2 ± 1.5	2
chondroitin sulfate	13.6 ± 4.8	2	4.1 ± 1.4	2
dermatan sulfate	7.1 ± 2.6	2	5.0 ± 1.5	2
hyaluronic acid	11.8 ± 4.1	5	5.6 ± 1.1	5

Human ADM was pretreated for 24 hr with the compounds shown above, washed in buffer, human keratinocytes were plated onto them in 24-well plates (100,000/ well), and the cultures continued for 3 or 10 days in KGM.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts from the indicated number (n) of experiments.  Means ± SEM.

TABLE 4

Keratinocyte Attachment and Growth on Antibiotic-Sterilized ADM

Effect of Attachment Agents

Treatment	Human keratinocytes/ mm of ADM cross-section
ADM without azide	3 day cultures	n	10 day cultures	n
control, untreated	7.8 ± 1.7	9	2.0 ± 0.7	9
fibrin	2.6 ± 1.5	2	1.8 ± 0.5	2
FCS - 2%	16.2 ± 6.3	7	19.6 ± 5.1 (p=0.001)	7
FCS - 20%	10.9 ± 4.3	6	23.5 ± 5.9 (p=0.001)	6
FCS - 100%	10.2 ± 3.2	2	18.5 ± 4.8	2
polylysine - 0.001%	7.7 ± 1.6	5	8.9 ± 1.8	5
heparin	4.7 ± 2.5	2	4.0 ± 2.0	2
hu Fb conditioned medium	13.6 ± 5.4	7	18.7 ± 5.3 (p=0.003)	7
co-cultured 3T3 cells	14.1 ± 4.8	7	15.6 ± 5.7 (p=0.04)	7

Human ADM was pretreated for 24 hr with the compounds shown above, washed in buffer, human keratinocytes were plated onto them in 24-well plates (100,000/ well), and the cultures continued for 3 or 10 days in KGM.  Alternatively, keratinocytes were maintained in the presence of 10% human fibroblast (hu Fb) conditioned medium, 2% FCS, or co-cultured 3T3 cells for the entire 3 or 10 day period.  ADMs were cryosectioned and adherent cells counted in twenty 40X fields.  Data are expressed as means of cell counts from the indicated number (n) of experiments.  Means ± SEM.

FYI.  The content on this blog is copyright protected by the author.  Feel free to read, copy, and disseminate the studies described here, but please indicate the origin of the work by citing this blog URL as an electronic web citation when it is appropriate to recognize and attribute the work of others.