Human keratinocyte isolation and cell culture: a survey of current practices in the UK

Btms Vol. 22, No. 1, pp. 35-39, 1996 Copyright @ 1995 Elsevier Science Ltd for ISBI Printed in Great Britain. Ail rights reserved 0305-4179/96 $15.00 + 0.00 ELSEVIER

0305-4179(95)00085-2

Human keratinocyte a survey of current J. T. Danielsl,

J. N. Kearney’

isolation practices

and cell culture: in the UK

and E. Ingham’

‘Yorkshire Regional Tissue Bank, Pinderfields General Hospital, Aberford Microbiology, The University of Leeds, UK

A survey was conducted to establish current techniques for isolation and culture of human keratinocytes. A questionnaire wassent to all units thought to be involved in keratinocyte culture, a total of 34 individuals; 62 per cent of those surveyed responded to the questionnaires. Tke proportion of individuals using high-calcium medium to culture keratinocytes was 53 per cent, while 47 per cent used low-calcium serum-free medium. The majority of replies followed trends dependent on the culture method employed. Details of anatomical donor skin site, keratinocyte isolation and culture were compared. In particular, the problems associafed wifk the use of commercially prepared low-calcium, serum-free medium were reported. The basic principles of keratinocyte culture reported in the literature were seen amongst all the replies received. It is interesting to note the variations in methods adopted as techniques are passed on and continuously modified to suif the requirements of the individual worker. This survey also highlights the difficulties that can occur when using massproduced complex media.

Bums, Vol. 22, No. I, 35-39,

1996

Road, Wakefield and ZDepartment

of

MCDB 151 was formulated which contained hydrocortisone and only I mg/ml fetal bovine serum proteins. The molarity of calcium in the recipe was 0.03 mM which stimulated keratinocyte colonies to grow as monolayers instead of entering the events of terminal differentiation. MCDB 152 was derived from MCDB 151 by the addition of various factors, such as EGF, ethanolamine, phosphoethanolamine and trace elements9.A chemically defined serum-free clonal culture medium, MCDB 153, was eventually developed“‘. The purpose of this questionnaire was to review current cell-culture practices in the UK, the aim being to compare and contrast methodologies and successrates. The work of our group involves the culture of human keratinocytes in low-calcium, serum-freemedium. After initial successwith this method, various problems have arisen without any changes in culture technique or product suppliers. This prompted the inclusion of specific questions in the survey about commercially prepared serum-free keratinocyte medium, with the aim of establishing whether our problems were isolated.

Introduction

Materials

The chosenmethod of keratinocyte culture dependson the final product required. For example, this may be a culture of terminally differentiating cells which will eventually form an epithelium suitable for grafting’, or a culture of activated, migrating keratinocytes similar to those found in a wound environmentz. Human epidermal cellshave been successfullygrown in serialculture to form keratinizing colonies3.A feeder layer of lethally irradiated 3T3 fibroblasts was required to initiate keratinocyte colony formation. This technique allowed viable, fibroblast-free keratinocyte clones to be isolated. Additional supplements, such as epidermal growth factor (EGF) and cholera toxin, enhancedproliferation, increased the lifespan of the keratinocyte cultures and enableda higher proportion of smallproliferating cells to be maintained4”. The requirement of a 3T3 feeder layer for the initiation and growth of keratinocyte cultures was eliminated by using medium 199 supplemented with pituitary extract fractions, hydrocortisone and whole bovine serum7. To move towards a more defined keratinocyte medium,

A questionnaire with an accompanying letter and stamped return envelope was posted to 34 individuals involved in keratinocyte cell culture. The questions included in the survey are listed in the Appendix.

and methods

Results Response to survey At 12 weeks from the posting date, 21 out of the 34 individuals surveyed (62 per cent) had responded. Two of these were not culturing keratinocytes at the time of the survey. Keratinocyte culture method Fifty-three per cent of the responding individuals used the high-calcium-medium method3 and 47 per cent employed the low-calcium, serum-freemethodlo. The majority of the replies received followed trends dependent on the culture method used. The results will be expressed for highcalcium-mediumusers(high Cat + ).

Burns: Vol. 22, No. 1, 1996

36

Donor skin sample Anatomical donor skin site High CaZ+ : The proportion of individuals using foreskin, breast, abdominal and skin from other sites was approximately 25 per cent for each. Low Ca2+ : Foreskin was the most popular choice (31 per cent) followed by breast (26 per cent), abdomen (26 per cent) and other sites (17 per cent). Average skin donor age High Ca2+ : The majority of individuals used skin samples from donors of all ages (59 per cent), 30 per cent from donors aged 21-50 years and 20 per cent from donors under 20 years. Low Ca2+ : Most individuals used skin samples from young donors aged o-20 years (47 per cent), 33 per cent from donors aged 21-50 years, and 13 per cent from donors > 50 years. Only 7 per cent useda variety of donor ages. Keratinocyte isolation The various conditions for epidermal removal and keratinocyte isolation are shown in Table1 for high-calciumTable I. Epidermal removal and keratinocyte described by individuals using high-calcium medium Method of epidermal keratinocyte isolation

removal

2.0-2.5 mg/ml dispase followed by 0.1-0.2594 EDTA for 1 O-20 min 0.25%

trypsin

O.l-0.25%

at 4’C overnight

protease XIV at 4°C overnight by trypsin/ETDA at 37°C for 20 min

of epidermal

25 U/ml

0.1%

isolation described by individuals

0.05%

using

using

% users of each method

isolation or without

0.02%

A wide range of feeder Feeder layer seeding density layer seeding densities was employed (1.3 x 104/cmZ to confluent). The most common seeding densi?y was 2.7 x 104/cmZ for both high- (50 per cent) and low- (,67per cent) calcium-mediumusers. Method of rendering 3T3s non-proliferative Approximately 70 per cent of all those who responded to the survey used y-irradiation and 30 per cent used 4-5 jtg/ml mitomycin-C. seeding and culture

Attachment and spreading, colony appearance and first passage timings The times of keratinocyte attachment and spreading, colony appearanceand first passage are shown in Figures 1-3 for both high- and low-calciummedium users.

0 ; 50 H x

33.3

0 or without

0.02%

at 37°C for 15 min

1 : 25 000 trypsin

Feeder layer strain High Ca’ ’ : 50 per cent usedSwissmouse3T3 strain J2, 10 per cent used EACC 88031146 and 40 per cent did noi know the origin of their 3T3 cells. Low Ca’ ’ : 67 per cent usedSwissmouse3T3 strain J2 and 37 per cent were uncertain of the origin of their 3T3 cells.

C

at 37°C for 10 min

1 : 250 trypsin

Use of a feeder layer High Caz+ : 80 per cent of responding individuals used a 3T3 feeder layer to initiate keratinocyte cell culture, 20 per cent occasionally used a feeder layer. Law Caz+ : 67 per cent of individuals always used a feeder layer, 33 per cent did not.

78

at 4°C overnight

0.05% 1 : 2500 trypsin with EDTA at 37°C for 20 min trypsin

at room

11

0.25% 1 : 250 trypsin with EDTA at 37’C for 30 min

0.25%

by individuals

11

of keratinocyte

Feeder layer

Keratinocyte seeding density A wide range of keratinocyte seeding densities was used (8.0 x 103/cm2 to 3.0 x 105/cmL). The most common seeding density was 2.7 x 10*/cmz for both high- (30 per cent) and low- (45 per cent) calcium-medium users.

% users of each method --

overnight

Table III. Keratinocyte low-calcium medium

Method

30 20 20

at 4°C overnight

dispase dispase

described

removal

2.0-2.5 mg/ml dispase temperature or 4°C 5 mg/ml

40

Cell yield The average cell yield was 3 x Io6/cmL ot donor skm

Keratinocyte h

TableII. Epidermal removal low-calcium medium

Method

% users of each method

at 4°C overnight trypsin with or without

at 37°C for l-3 trypsin

10 mg/ml followed

and

isolation

medium users and in Tables11 and II[ for low-calcium. medium users.

at 25°C for 5-l 0 min

33.3 11 .l 11 .I II .I

2-3

Keratlnocyte

24

48

(hours) attachment

72,

3-7 +------A and

14

(days) spreading

Figure 1. The percentageof individualsreporting the time, in hours or days,of keratinocyte attachmentandspread@, in hi& ( W) and low Cl) calciummedium.

Daniels

et al.: Human keratinocyte

isolation

and cell culture

37

Problems with medium Every person who replied to the survey had experienced one or more problems with commercially available serum-free keratinocyte medium. These problems included slow growth rates (30 per cent), granulation inside the keratinocytes (26 per cent), lengthy primary culture time, i.e. > 2 weeks (22 per cent), above average failure of cultures to ‘take’ (18 per cent) or other problems (4 per cent). Batch variation in medium purchased Forty-four per cent of those purchasing serum-free medium had not experienced batch-to-batch variation in medium performance,33 per cent had experienced variation and 24 per cent could not comment.

I7

10

14

Figure 2. The percentageof individualsreporting the time, in

In-house medium preparation All the individuals using low-calcium, serum-freeculture had not attempted to make their own medium.

days,of the first appearance of coloniesin high ( n ) andlow (0) calciummedium.

Discussion

First

colonies

(days)

100 m m .-G g 50 C H x

C First

11-14

15-21

passage

time

>22 (days)

Figure 3. The percentageof individualsreporting the time, in

days, of the first keratinocyte culture passagein high ( n ) and low ( q ) calciummedium.

Replenishment of 3T3s High CaZ+ : 50 per cent of those who responded added fresh 3T3s to keratinocyte cultures, 40 per cent did so occasionally and 10 per cent never did. Low CaZ’ : 100 per cent did not add fresh 3T3s to keratinocyte cultures. Ratio of primary culture expansion High Ca2+.. The expans’o I n ratio range was I : 2 to 1: 10, the most common expansion ratio was I : 5 (30 per cent). Low Ca2+ . The expan sion ratio ranged from I : I to I : 10. Serum-free culture medium Several questions were askedspecifically about serum-free culture medium. Medium supplier Of those individuals culturing keratinocytes in low-calcium, serum-free conditions, 90 per cent purchasedmedium from Gibco Life Technologies and 10 per cent from Sigma, and all used the culture supplements recommended/supplied by each company.

The ability to culture human keratinocytes is of great importance both to the clinician and to researchworkers. The aim of this study was to identify the current keratinocyte culture practices in the UK and to highlight any common difficulties. The questionnaire gave rise to a variety of answers which indicated deviations from original published protocols. Of the individuals surveyed, 53 per cent used adaptations of the high-calcium-medium culture method (high Ca2+)3 and 47 per cent used methods similar to the low-calcium, serum-free system (low CaZ+ )I“. The majority of individuals culturing keratinocytes for clinical purposes tended to use the high-calcium method. This trend may reflect a reluctance to use bovine pituitary extract (one of the supplements required in serum-free culture), since this protein can be very expensive when purchased from a source of guaranteed BSE-free cattle, compared with screenedfetal calf serum. No real preference with regard to skin sampledonor site or donor age, from which to isolate keratinocytes, was expressedby high Ca” + users,despite one report suggesting that plating efficiency and culture lifespanarereduced in keratinocytes from older donors3. Those individuals using low CaZ+ generally used foreskin from donors under 20 years. For keratinocyte isolation, the high CaZf respondents used a variety of treatments which all involved trypsin in either a one- or two-step process,while all low CaZ+ respondents used dispase enzyme overnight to remove the epidermis, followed by a trypsinization step. The average cell yield from the processeswas 3 x 106/cm2 of donor skin. The majority of individuals regularly used a 3T3 fibroblast feeder layer (high Ca2+ 80 per cent, low Ca”+ 67 per cent), which has usually been y-irradiated (70 per cent) and seededat 2.7 x 104/cmZ(high Ca’+ 50 per cent and low Ca2+ 67 per cent). Isolation of keratinocytes by the Rheinwald and Green method3 involves the trypsinisation of a whole skin piece yielding a variety of cell types, including dermal fibroblasts. As well as serving as an anchor for keratinocyte attachment, a growth-arrested 3T3 feeder layer suppressesthe growth of dermal fibroblasts, hence most individuals using the high-Ca2+ method used a feeder layer. The enzyme dispase, used by all the low-Caz + users to remove the epidermis, acts at the dermal-epidermal junction. This yields an epidermal sheet

Burns: Vol. 22, No. 1, 1936

38

in which keratinocytes are the majority cell type. However, even when using dispase to remove the epidermis and trypsinization to isolate the keratinocytes, 67 per cent of low-Caz + medium users still required a feeder layer to initiate keratinocyte culture. This is contrary to the requirements of culture initiation in MCDB 153 reported by Boyce and HamI“, i.e. that a feeder layer was not necessary. In our experience, a feeder layer is essential when culturing keratinocytes, isolated from epidermis removed from foreskin with dispase, to prevent dermal fibroblast overgrowth in serum-free medium. Although a wide variety of keratinocyte seeding densities was reported, 2.7 x 104/cmZwas the most common for both culture methods. This density gave a 1: 1 keratinocyte : 3T3 feeder cell ratio in most cases.Those individuals using high CaZ+ generally reported keratinocyte attachment and spreading within 24 h (60 per cent), first colony appearance at 3 days (40 per cent) and first passagewithin 2 weeks (90 per cent). Those using low CaZ+ observed much greater variation in keratinocyte progress, i.e. attachment and spreading between 2 h and 2 weeks, first colony appearanceat 5 days (44.4 per cent) and first passagealso within 2 weeks (66.6 per cent). Although responsesto the questions about cell attachment and first colony appearance were probably subjective they were considered to warrant inclusion in the questionnaire analysisto give an idea of the time-scaleover which these events are likely to occur in culture. If different keratinocyte strains continually require excessively long culture periods to achieve attachment and colony formation, these parameters would be among those which would alert attention to a problem with the culture technique or medium. Indeed, the variability in responses of those individuals culturing keratinocytes with low-calcium medium compared with those culturing in high-calcium medium indicated that the former system may be more susceptibleto alterations in culture protocols. Only 10 per cent of individuals using high Caz+ never replenished the 3T3 cells in their keratinocyte cultures, in contrast to 100 per cent of those using low CaZ+ Questions were included in the survey which reflected the problems this laboratory had experienced with commercially available serum-free keratinocyte medium. The majority of individuals using serum-free culture purchased their medium from Gibco (90 per cent) and 100 per cent had experienced one or more problems. These problems included slow growth rates (30 per cent), granulation inside the keratinocytes (26 per cent), lengthy primary culture times, i.e. > 2 weeks (22 per cent), above-average failure of cultures to ‘take’ (18 per cent) or other problems (4 per cent). Of th ose purchasing serum-free medium, 33 per cent had experienced batch-to-batch variation in medium. Despite the reported successof keratinocyte culture in serum-freemediumlO, the wide range in parametersreported in this survey, such as attachment and spreading, first colony appearance, first passagetime and various problems, demonstrated the variability in’ culture progress experienced amongst individuals. The fact that the useof a feeder layer is still favoured for serum-freeculture suggests that manufactured serum-free media may not perform as well without 3T3s asdid the original recipesfrom which it was developed7-‘0. The preparation of a low-calcium, serum-free keratinocyte culture medium, MCDB 153, has been describedll. The preparation of stock solutions and conditions for final medium preparation were clearly

defined. It is possible that in scaling up of similar procedures to a commercial scale,it is more difficult to produce the medium to precisely the optimal recipe. Conversely, the range of protocol procedures employed by users OS‘ high-Ca2+ medium did not seemto affect culture success, as demonstrated by the smaller ranges in the aforementioned parameters and the absenceof reports or particular problems. This system appears to be more resiliant to alterations in culture procotols than the serumfree method. This survey has attempted to give an overview oi current practices in the UK and to highlight the differences between culture methods. The culture method adopted should be selected on the basis of the nature of the keratinocyte population required, i.e. proliferating or ter. minally differentiating. As a result of this survey, it was found that the problems experienced with serum-fret* culture in this laboratory have not been isolated. This prompted further work which has identified and resolved some of the problems so far experienced with commercially prepared, low-calcium, serum-free keratinocyte cuiture mediumlz. The authors recommend the development of an optimized protocol of keratinocyte culture for j particular manufactured serum-free medium, This should reduce the variation in culture performance such as that described in this survey.

Acknowledgements The authors are grateful to everyone who was involved in compiling this survey and to Derren Waghorn for assistance with the manuscript.

References 1 GreenH, Kehinde0, ThomasJ. Growth of culturedhuman epidemalcellsinto multipleepitheliumsuitablefor grafting. Proc Nutl Acad Sci USA 1979; 76: 5665-5668.

2 Grinnell F. The activated keratinocyte: up re&ation of celi adhesionand migration during wound healing. J Trauma 1990; 30: s144-s149.

3 RheinwaldJG,GreenH. Serialcultivation of strainsof human epidermalkeratinocytes:the formation of keratinisingcoloniesfrom singlecells.Cell 1975; 6: 331-344. 4 Cohen S. The stimulationof epidermalproliferation by a specificprotein (EGF).Dat Viol 1965;12: 394-407. 5 RheinwaldRG, Green H. Epidermalgrowth factor and the multiplication of cultured human epidermalkeratinocytes. Nature 1977;265: 4.21-424. 6 Green H. Cyclic AMP in relation to proliferation of the epidermalcell: a new view. Cell 1978;15: 801-811. 7 PeehlDM, Ham RG. Growth and differentiation of human keratinocyteswithout a feederlayer or conditionedmedium. In Vifro 1980; 16: 516-525.

8 PeehlDM, HamRG. Clonalgrowth of humankeratinocytes with smallamountsof dialysed serum.In vitro 1980; 16: 526-538. 9 TsaoMC, Walthall BJ,Ham RG. Clonal growth of normal

human keratinocytes in a defined medium.] Cell _Physid 1982;110: 219-229. 10 Boyce ST, Ham RG. Calcium regulated di&rentiation chemically defined donal culture and serum-free serial culture.] InvestDemfof 1983; 81: ~33~40.

Daniels

et al.: Human

keratinocyte

isolation

and cell culture

Boyce ST, Ham RG. Normal human epidermal keratinocytes. In: Webber MM, Sekely L, eds. In Vifro Models for Cancer Research. Boca Raton, FL: CRC Press, 1985; pp 245-273. 12 Daniels JT, Harris IR, Keamey JN, Ingham E. Calcium: a crucial consideration in serum-free keratinocyte culture. Erp Dermafol 1995; 4: 183-191. 11

Appendix 1. From which anatomical site do you obtain skin samples for keratinocyte isolation? 2. What is the average donor age? 3. How do you remove the epidermis from the skin sample? 4. In what medium is the de-epidermalizing enzyme used? 5. What is the concentration of the enzyme? 6. What are the incubation conditions for de-epidermalization? 7. Which trypsin do you use to disaggregate the keratinocytes from the epidermis? 8. At what concentration is the enzyme used? 9. For how long and at what temperature do you trypsinize the epidermis? 10. What is the average cell yield per cm27 11. Do you use a 3T3 feeder layer? 12. If yes, which strain? 13. If yes, at what density is the feeder layer seeded?

39 14. How are the 3T3s rendered non-proliferative? 15. At what density do you seed freshly isolated keratinocytes? 16. How quickly do keratinocytes attach and spread? 17. When do the first colonies appear? 18. How soon can the culture usually be passed? 19. Are fresh 3T3s added to the culture? 20. Into how many flasks may you expect to be able to expand a primary culture from a 75-cm’ flask at 70-80 per cent confluence? 21. From whom do you purchase serum-free medium? 22. Which additives do you use? 23. Have you experienced problems with the medium you have purchased? 24. Do you experience batch-to-batch variation with purchased serum-free medium? 25. Have you attempted to make your own serum-free medium?

Paper accepted

after revision

27 May

1995.

Correspondence should be addressed to: JT Daniels, Research Assistant/PhD student, Regional Tissue Bank, Pinderfields Hospitals Trust, Wakefield WFl 4DG, West Yorkshire, UK.