Colony Growth of Human Hematopoietic Progenitor ...

Published October 1, 1996

C o l o n y G r o w t h o f H u m a n H e m a t o p o i e t i c Progenitor Cells in the A b s e n c e o f S e r u m Is S u p p o r t e d by a Proteinase Inhibitor Identified as A n t i l e u k o p r o t e i n a s e By Henri~tte M. Goselink,* Jo van Damme,~ Pieter S. Hiemstra,:~ Anja Wuyts,~Jan Stolk,:~Willem E. Fibbe,* Rod Willemze,* and J.H. Frederik Falkenburg* From the *Laboratory of Experimental Hematology, Department of Hematology and *the Department of Pulmonology, Leiden University Hospital, Leiden, 2300 R C, The Netherlands; and the ~Rega Institutefor Medical Research, University qf Leuven, Leuven, Belgium

Summary

haracterization of hematopoiesis in humans is dependent on the in vitro analysis of hematopoietic progenitor cells (HPC) 1. HPCs can be studied using semisolid medium cultures, and their potential for proliferation and differentiation can be characterized by their ability to form

C

IAbbreviationsusedin thispaper:oe-1 PI, or-1 proteinaseinhibitor; ALP, antileukoproteinase; bFGF, basic fibroblast growth factor; BFU-E, burstforming unit erythroid colony; CM, conditioned medium; CPG, controlled pore glass; EGF, epidermal growth factor; EPO, erythropoietin; GCSF, granulocyte colony stimulating factor; GM, granulocyte macrophage; HGF, hematopoieticgrowth factor; HPC, hematopoieticprogenitor cell; HSA, human serum albumin; huALP, human antileukoproteinase; IGF, insulin-like growth factor; PDGF, platelet-derived growth factor; rALP, recombinant human antileukoproteinase; RP-HPLC, reverse-phaseHPLC; SCF, stem cell factor; TFA, trifluroaceticacid. 1305

colonies of hematopoietic cells in these cultures. The growth and differentiation of H P C in colony assays depend on the presence of hematopoietic growth factors (HGF) in the cultures (1). These H G F may be added to the culture medium or produced by accessory cells or H P C as paracrine or autocrine growth factors, respectively. In addition, factors that are present in serum, including insulin, cholesterol, or albumin, are critical for the in vitro growth of H P C (2-7). BSA, frequently used as a source of albumin, is usually not >95-99% pure because of its strong proteinbinding properties (8, 9). Since in serum-free colony assays, BSA is often used at relatively high concentrations, residual undefined proteins may significantly contribute to the growth-supportive potential of BSA. Pilot studies in our laboratory indicated that in contrast to these BSA preparations, highly purified concentrates of clinical-grade human albumin did not support the proliferation of H P C from

j. Exp. Med. 9 The Rockefeller University Press ~ 0022-1007/96/10/1305/08 Volume 184 October 1996 1305-1312

$2.00

Downloaded from jem.rupress.org on June 4, 2013

Serum contains many growth factors and nutrients that stimulate colony formation of hematopoietic progenitor cells (HPC) in semisolid cultures, In the absence of serum, no proliferation of HPCs could be obtained in semisolid medium cultures of partially purified bone marrow cells in the presence of multiple hematopoietic growth factors, insulin, cholesterol, and purified clinical-grade human albumin. This appeared to be due to a suppressive activity induced by monocyte- and T lymphocyte-depleted accessory cells on CD34 + HPCs. Serum-free conditioned medium from the bladder carcinoma cellline 5637 could replace serum to support the growth of HPCs in these cultures. After gel filtration and reverse-phase high-performance liquid chromatography of 5637 supernatants, this activity could be attributed to a 15-kD protein that was further identified by NH2-terminal sequence analysis as the serine proteinase inhibitor antileukoproteinase (ALP). The growth-supportive activity from the 5637 conditioned medium and the (partially) purified fractions could be completely neutralized by a polyclonal rabbit IgG antibody against human ALP (huALP). Similar supportive effects on the growth of H P C could be obtained in the presence of recombinant huALP. We demonstrated that the COOH-terminal domain of ALP containing the proteinase inhibitory activity was responsible for this effect, o~-1 proteinase inhibitor was capable of similar support of in vitro H P C growth. These results illustrate that proteinase inhibitors play an important role in the in vitro growth of hematopoietic ceils by the neutralization of proteinases produced by bone marrow accessory cells. This may be of particular relevance for in vitro expansion of human hematopoietic stem cells in serum-free media.


m o n o c y t e - and T l y m p h o c y t e - d e p l e t e d mononuclear bone marrow cells in semisolid m e d i u m cultures in the presence o f H G F , cholesterol, and insulin. In this study, we analyzed the mechanisms o f this finding and showed that this was due to a suppressive effect o f m o n o c y t e - and T l y m p h o c y t e - d e p l e t e d bone marrow accessory cells on C D 3 4 - p u r i fled H P C . W e investigated which serum factor, in addition to human albumin, insulin, and cholesterol, was responsible for supporting the in vitro proliferation o f bone m a r r o w derived H P C . W e found that a 15-kD protein present in conditioned m e d i u m (CM) from the bladder carcinoma cell line 5637 could replace serum. This protein was identified as the serine proteinase inhibitor antileukoproteinase (ALP).

Materials and M e t h o d s Cell Preparations. After informed consent, bone marrow sam-

1306

Antileukoproteinase Supports In Vitro Hematopoiesis


ples were obtained from healthy donors. The cells were centrifuged on Ficoll Isopaque (density 1.077 g/cm 3, 20 min, 1,000 g), and mononuclear cells were depleted of monocytes and T lymphocytes as described previously (10). These enriched bone marrow samples were cryopreserved and stored in liquid nitrogen. hnmediately before use, the cells were thawed rapidly in a water bath of 37~ washed once in cold (0~ RPMI-1640 medium (Hyclone Labs., Logan, UT) supplemented with 20% (vol/vol) fetal bovine serum (Boehringer, Mannheim, Germany), and twice with cold RPMI-1640 medium supplemented with 0.1% (wt/vol) clinical grade human serum albumin (HSA; Central Laboratory of the Bloodtransfusion Services, Amsterdam, The Netherlands). The cells were resuspended in IMDM (Hyclone Labs.) supplemented with 0.1% (wt/vol) HSA. In some experiments, highly purified CD34 + cells were isolated from the monocyteand T lymphocyte-depleted bone marrow cells. The CD34 + cells were either selected after staining with FITC-conjugated CD34specific mAbs (HPCA-2; Becton Dickinson, Mountain View, CA) and sorted with a FACStar| PLUS (Becton Dickinson) or isolated after incubation with magnetic Dynabeads M-450 (Dynal Inc., Lake Success, NY) coated with a primary antibody specific for the CD34 Ag. After magnetizing the CD34 + cells, the nonbinding population was harvested as the CD34 accessory cells. The CD34 + cells were recovered after incubation with a sheep polyclonal antibody anti-CD34 Fab IgG (Detachabead CD34; Dynal Inc.), resulting in >80% pure CD34 + cells. Cytokines and Antibodies. All cytokines used in this study were recombinant human proteins and used at plateau concentration to induce optimal colony growth. The cytokines and final concentrations used were: stem cell factor (SCF) (50 ng/ml), granulocyte-colony stinmlating factor (G-CSF) (10 ng/ml), basic fibroblast growth factor (bFGF) (10 ng/ml), and epidermal growth factor (EGF) (10 ng/ml) provided by Amgen (Thousand Oaks, CA); IL-3 (50 ng/ml), IL-11 (10 ng/ml), monocyte-CSF (1,000 U/ml), and GM-CSF (10 ng/ml) were provided by Genetics Institute (Cambridge, MA); IL-lc~ (10 ng/ml), IL-5 (10 ng/ml), and IL-12 (10 ng/ml) were gifts from Hoffman La Roche (Nutley, NJ); IL-4 (500 U/ml) and IL-7 (10 ng/ml) were gifts from hnmunex Corporation (Seattle, WA); IL-6 (1 ng/ml) and IL-8 (10 ng/ml) were gifts from Sandoz (Basel, Switzerland); plateletderived growth factor (PDGF; 30 ng/ml) was obtained from GIBCO BRL (Gaithersburg, MD); insulin-like growth factors (IGF-I, IGF-II, 250 ng/ml) were obtained from Genzyme Cor-

poration (Cambridge, MA); and erythropoietin (EPO) (1 1U/ml) was a gift from Cilag (Herentals, Belgium). IL-9 (1,000 U/ml) was provided by Ludwig Institute for Cancer, (Brussels, Belgium), IL-10 (10 ng/ml) was obtained from Schering-Plough Research Institute (Kenilworth, NJ), and IL-2 (300 U/ml) was a gift from Roussel Uclaf (Paris, France). Recombinant human antileukoproteinase (rALP) was a generous gift from Dr. R.C. Thompson (Synergen Inc., Boulder, CO). The separated NHa-tenninal and the COOH-terminal domains of ALP, designated ALP-D-1 and ALP-D-2, respectively, were kindly donated by I)r. G. Steffens (Grtinentahl, Aachen, Germany) and were used at equivalent molar concentrations ofrALP, er proteinase inhibitor (0r PI), used at a final concentration of 0.2 to 20 Ixg/ml, was obtained from Cutter Biologic (Hartford, CT). A neutralizing anti-human antileukoproteinase (huALP) antibody was prepared in rabbits. The antibody has been found to have a high titer and specificity for huALP (11). Clonogenic Assay. Semisolid medium cultures were perfomaed in sixfold in flat-bottomed 96-well microtiterplates (Greiner, A1phen a/d Rijn, The Netherlands) in aliquots of 0.1 ml per well, containing 103 monocyte- and T lymphocyte--depleted mononuclear bone marrow cells or 150 purified CD34 § cells per well. The culture medimn consisted of IMDM, supplemented with 0.6% (wt/vol) HSA, 20 Ixg/ml cholesterol (Sigma Chemical Company, St. Louis, MO; C-7402), 10 txg/ml insulin (Signna 1-4011), 5 X 10 `5 M ~-mercaptoethanol, (3.47 mg/ml human transferrin (Behringwerke AG, Marburg, Germany) saturated with FeCI 3 9 6H20, methylcellulose (Methocel 4000 cps; Fluka, Freiburg, Germany) at a final concentration of 1.l% (wt/vol) and a mixture of cytokines: GM-CSF (10 ng/ml), G-CSF (10 ng/ml), IL-3 (50 ng/ml), SCF (50 ng/ml), and EPO (1 IU/ml). All cytokines were present at saturating concentrations. This culture medium was supplemented either with control medium, hunran AB serum from healthy blood donors (10% [vol/vol]) that was prescreened for alloantibodies and inactivated for 30 rain at 56~ or a CM produced by the human bladder carcinoma cell line 5637 in the absence of serum, or (partially) purified fractions from this CM or proteinase inhibitors. Purified CD34 + cells were cultured in the absence or presence of increasing numbers ofmonocyte- and T lymphocyte-depleted bone marrow accessory cells. In some experiments, to neutralize the biologic activity of these supplements, a 2-h incubation at 37~ and 5% CO2 in the presence of neutralizing concentrations ofanti-huALP antibodies was performed before initiation of the bone marrow culture. After 12 d of incubation in a fully humified atmosphere of 5% CO 2 at 37~ colonies were counted. Granulocyte-macrophage colony-forming units (CFU-GM) were defined as aggregates of at least 50 cells and burst-forming unit erythroid colonies (BFU-E) as henmglobinized bursts. To calculate the HPC growth in the partially purified bone marrow fractions, the colony growth induced by the culture medium alone was considered background and therefore in these experiments subtracted from the colony growth induced by AB serum, CM, or purified CM fraction. BFU-E and CFU-GM growth in the presence of 10% (vol/vol) AB serum was considered 100% growth (control culture). Colony growth was expressed as percentage growth of the control culture. Production of CM. The human bladder carcinoma cell line 5637 (12, 13) was used to produce a serum-free CM. Confluent adherent cells were washed three times with RPMI-1640 and incubated for 24 h in 1KPMI-1640 serum-free medium, supplemented with 0.6% (wt/vol) HSA. Then, the cells were washed three times with RPMI-1640 and cultured in IMDM, supplemented with 5 X 10 -s M ~-mercaptoethanol and 0.47 mg/ml


-AB serum

Active P-,P-HPLC fractions, containing low protein concentrations, were analyzed for purity by SDS-PAGE. Samples (20 p,1) were loaded onto a polyacrylamide gel (reducing conditions), run simultaneously with molecular mass markers, and proteins were silver stained. The electrophoretically pure fractions were finally analyzed on a 477A protein sequencer with on-line detection of the phenylthiohydantoin AA in a 120A analyzer (Applied Biosysterns) (17). Chemical digestion of protein was obtained with 75% formic acid at 37~ for 50 h (17).

+AB serum

0:1 2:1 O

4:1

+

16:1

8:1 CFU-GM

(o

c)

o:I 2:1 4:1

I

Results

8:1

Influence of Serum and Accessory Cells on H P C Growth.

BFU-E

16:1 150

100

50

0

50

100

150

p e r c e n t a g e c o l o n y g r o w t h of control

Figure 1.

human transferrin, saturated with FeCIO 3 9 6H20. After 48 h, the culture supematant was harvested. The crude supernatant was centrifuged (15 rain, 500 g, 20~ irradiated (50 Gy), and frozen at - 2 0 ~ after addition of 0.1% (wt/vol) HSA for stabilization. Purification of CM. Crude supernatant from 5637 cells was incubated for 2 h at 4~ with controlled pore glass (CPG) beads (PG-350-200; Sigma) for adsorbtion at pH 7.0 (14-16). The CPG beads were washed with Dulbecco's modified PBS and 0.01 M glycine/HC1, pH 3.5. The protein fraction was eluted by carefully stirring with 0.3 M glycine/HC1, pH 2.0 (1/10 of original crude volume), concentrated and neutralized by dialysis against PBS containing 10% (wt/vol) polyethylene glycol (PEG 20.000) (Serva, Heidelberg, Germany). For further fractionation of proteins, a gel filtration was performed on Ultrogel ACA 54 (Pharmacia, Uppsala, Sweden) by using a 2.6 • 100-cm column and a flow rate of 22 ml/h. BSA (Mr 67,000), OVA (M r 45,000), chymotrypsinogen (M r 25,000), and lysozyme (M r 14,300) were used as molecular mass markers. 5.5-ml fractions were collected by elution with 18% ethylene glycol buffer in 1.55 M NaCI, 8 m M phosphate, pH 7.2. The protein concentration was analyzed by a Coomassie blue/G-250 binding assay, using the Bio-Rad (Richmond, CA) commercial kit and BSA as a standard. Fractions from the Ultrogel column that could support colony growth of HPC were supplemented with Tween 20 (0.01% vol/vol), concentrated on a centricon-3 filter (Amicon Inc., Beverly, MA), and further fractionated by reverse-phase HPLC (P,P-HPLC). The concentrate was injected on a C8-Aquapore ILP-300 (2.1 • 220 ram) column (Applied Biosystems Inc., Foster City, CA) and equilibrated with 0.1% trifluoroacetic acid (TFA). Proteins were eluted at a flow rate of 0.4 ml/nfin with a linear gradient of acetonitrile (0-80% in 80 min) in 0.1% TFA. 400-p~l fractions were collected in tubes containing 0.01% (final concentration) Tween 20, and the absorption at 220 n m was measured. 1307

Goselink et al.

(,

=

8).

R e p l a c e m e n t o f AB serum b y H S A (0.6% w t / v o l ) , c h o lesterol (20 p~g/ml), and insulin (10 ~ g / m l ) resulted in only 4 _+ 1 G M colonies and n o B F U - E g r o w t h per 103 m o n o cyte- and T l y m p h o c y t e - d e p l e t e d m o n o n u c l e a r b o n e m a r r o w cells plated (background c o l o n y growth). In contrast, in cultures o f purified C D 3 4 + cells, H P C g r o w t h in the presence o f 10% AB serum yielded 16 --- 1 C F U - G M and 10 +- 1 B F U - E per 150 cells plated, whereas in the absence o f AB serum, 9 + 2 C F U - G M and 6 -+ 1 B F U - E were cultured (n = 10). Fig. 1 shows that addition o f m o n o c y t e and T l y m p h o c y t e - d e p l e t e d C D 3 4 - b o n e m a r r o w cells to purified C D 3 4 + ceils resulted in a d o s e - d e p e n d e n t i n h i b i tion o f H P C g r o w t h in the absence, b u t n o t in the presence o f AB serum, illustrating that serum contains factors capable o f supporting H P C g r o w t h in the presence o f b o n e m a r r o w accessory cells.

Influence of 5637 C M on H P C Growth in the Absence of Serum. R e p l a c e m e n t o f AB serum b y interleukins (IL-1 t h r o u g h [L-12), separately or in various c o m b i n a t i o n s , or g r o w t h factors P D G F , b F G F , E G F , I G F - I , I G F - I I , or m a c -

Effectof 5637 CM on Colony Growth of Monocyte- and T Lymphocyte-depleted Bone Marrow Mononuclear Cells in the Absence of Serum

T a b l e 1.

Colonies per 103 cells plated Condition Serum-free medium 5637 CM (vol/vol)

AB serum (vol/vol)

10% 3% 1% 10%

CFU-GM

BFU-E

4 + 1 28 + 3 18 -+ 3 9_+2 21 + 2

0 17 + 2 13 _+ 3 7_+2 20 +_ 2

All cultures were perfomled in the presence of optimal concentrations ofHGF (GM-CSF 10 ng/ml, G-CSF 10 ng/ml, IL-3 50 ng/ml, SCF 50 ng/ml, and EPO 1 IU/ml).


Influence of CD34- accessory cells on CFU-GM and BFU-E growth of purified CD34 + cells in the presence of multiple HGF (G-CSF 10 ng/ml, GM-CSF 10 ng/nfl, IL-3 50 ng/ml, SCF 50 ng/ml, EPO 1 IU/nfl) in the absence and presence of AB serum (10% vol/vol) (n = 2). Colony growth of 150 CD34 + cells plated per well, supplemented with increasing numbers of CD34- cells at ratios up to 16:1, is expressed as percentage of growth of control culture, obtained in the presence of 10% (vol/vol) AB serum in the absence of CD34- accessory cells. Absolute numbers of colonies in the control culture were 8 +-- 1 CFU-GM and 11 -+ 1 BFU-E (mean +- SE per 15(:)cells plated).

In cultures c o n t a i n i n g 10% AB serum and H G F (IL-3, G M - C S F , G - C S F , SCF, E P O ) , the absolute n u m b e r s o f H P C per 103 m o n o c y t e - and T lymphocyte--depleted b o n e m a r r o w cells were 21 + 2 C F U - G M and 20 + 2 B F U - E


160

140

/

,,J

o "

I'Z

120

o ,, 1 0 0 o -t-

o

--IP C F U - G M BFU-E

"

eo o ee

>-

6O

o o o

40

o

2o UJ

er

"' el

0

-20 26

I 28

I 30

~ 32

r 34

I 36

I 38

f 40

42

FRACTION

2. Fractionation of HPC growth-promoting activity from CPG-purified 5637 CM by gel filtration (Uhrogel ACA 54). The shaded area represents the protein concentration per fraction. The growth supportive activityfor CFU-GM (x-x) and BFU-E ([Z]-lZ])coloniesfrom the monocyte- and T-lymphocyte depleted bone marrow mononuclear cells was analyzedin the presence of multipleHGF in the absenceof serum 0l = 4). Colony g-rowth supported by each colunm fraction at a final concentration of0.3% was expressed as percentage of growth of control cultures in the presence of 10% AB serum. The absolute numbers of colonies in the control cuhure were 17 --+ 4 CFU-GM and 20 .+ 4 BFU-E (mean -+ SE per 1()~ cells plated). Protein concentrationswere determined by the Coomassie blue binding assay. Molecular mass markers: BSA, (M~ 67,I)00), ovalbmnin (OV, M~ 45,000), chymotrypsinogen (CHYM, M, 25,i)011), and lysozyme (LYS, M,. 14,300). Figure

rophage CSF, in addition to the cytokines present in the culture medium, did not result in additive growth of colonies from m o n o c y t e - and T lymphocyte-depleted bone marrow cells (data not shown). However, addition of 3 10% (vol/vol) CM, produced by 5637 cells in the absence of serum, could support the colony growth up to levels similar to the growth observed in the presence of 10% AB serum. Replacement of AB serum by 10% (vol/vol) 5637 C M to the culture m e d i u m supported 136 + 12% of C F U - G M growth and 85 + 4% of B F U - E growth as compared to the control culture. These results show that in the presence of purified hulnan albumin, 5637 cells produce factors that can support the growth of C F U - G M and BFU-E comparable to the effect of AB serum (Table 1).

Characterization of the Supportive Factor from 5637 CM. The protein fraction from the 5637 CM, concentrated by adsorption to and elution froln the C P G beads, contained the colony growth supportive Factor with a fivefold higher concentration of activity compared to the crude 5637 supernatant (data not shown). After gel filtration of the C P G 1308

Figure 3. Purification of HPC growth-promoting activity by RPHPLC. In the absence of serum, each P,P-HPLC fraction (400 b~l)was analyzedfor the supportiveactiviWff)r CFU-GM. (x-x) and BFU-E ([Z]-U]) growth from monocyte- and T lymphocyte-depleted bone marrow mononuclear cells in the presence ofHGF (tl = 3). Colony growth supported by each fi'actionat a final concentration of 0.3% was expressed as percentage of growth of control cultures, obtained in the presence of II)% (vol/vol) AB serum. Absolutenumbers of colonies in the control culture were 12 -+ 3 CFU-GM and 16 _+ 4 BFU-E (mean .+ SE per 103 cells plated). Fractionf-32 contained the highest level of supportive activity'for both the myeloidand erythroid colonygrowth, 75 .+ 12"/,and 121 ,+ 20% (mean percentage of growth ,+ SE), respectively.

eluate into 90 fractions, the major protein peaks were recovered from fractions 30 to 40 (>45 kD). As illustrated in Fig. 2, addition of0.3% (vol/vol) of these high M r proteincontaining fractions to the culture m e d i u m did not result in an enhancement of colony growth. However, addition of 0.3% (vol/vol) of low Mr protein fractions 59 through 72, containing very low protein concentration ( < 1 - 2 ~g/ml) supported the growth of both C F U - G M and BFU-E, with a maximal level of growth in fractions 62 and 69 containing 10-15-kD protein. These fractions were chosen for further purification by R P - H P L C technique. After HPLC separation of fraction 62, the enhancing activity for both G M colony growth and B F U - E growth was found in fractions 30-32 (Fig. 3). Fraction 32 supported C F U - G M and BFU-E growth up to 75 + 12% and 125 + 20% of control cultures, respectively. The HPLC fractions, which contained the highest growth supportive activity, were electrophoretically analyzed for purity by means of SDS-PAGE, and the molecular mass was defined as shown in Fig. 4. The HPLC fractions 30-32 from gel fihration fraction 62 showed a distinct protein band of 15 kD, with the highest concentration in fraction

Antileukoproteinase Supports 111Vitro Hematopoiesis


f

IM


32(g32). An automated protein sequencer was used to sequence the first 30 N H 2 R - t e r m i n a l amino acid residues o f the protein, and they revealed complete structural identity with h o m o l o g y to the NH2-terminal sequence o f the serine fraction

Influence of Antileukoproteinase on the Suppressive Effect of Accessory Cells on H P C Growth in the Absence of Serum. M o n o c y t e - and T l y m p h o c y t e - d e p l e t e d bone marrow cells were separated with Dynabeads into C D 3 4 - cells and a population containing 80% C D 3 4 + cells. Fig. 8 shows the colony growth o f these purified C D 3 4 + cells in the absence or presence o f increasing numbers o f C D 3 4 - cells. In two separate experiments, it was shown that the inhibited growth o f C F U - G M and B F U - E , induced by the C D 3 4 - accessory cells in the absence o f serum, was completely reconshtuted by the addition o f rALP (100 ng/ml) or C M (10%

sequence

(f)

HPLC f3a from gel f i l t r a t i o n f62 no d i g e s t i o n

1

5

i0

15

20

25

30

[

I

i

i

I

I

I

SGKSFKAGVCPPKKSAQCLRYKKPECQSDW

HPLC f30+31 from gel f i l t r a t i o n f62 HCOOH digestion

55

60

65

70

75

1

i

I

I

1

I

PVDTPNPTRRKPGKCPVTYGQCLMLNP

1 5 i0 15 20 25 30 35 40 45 1 [ I I I I [ I I i SGKSFKAGVCPPKKSAQCLRYKKPECQSDWQCPGKKRCCPDTCGI

HPLC f28 from gel f i l t r a t i o n f68+69 no d i g e s t i o n

antileukoproteinase

50

1

1

5

i0

15

20

25

30

35

40

45

50

55

60

65

70

75

80

L

I

I

I

1

L

I

L

I

I

I

h

I

L

I

L

I

SGKSFKAGVCPPKKSAQCLRYKKPECQSDWQCPGKKRCCPDTCGIKCLDPVDTPNPTRRKPGKCPVTYGQCLMLNPPNFC 81

85

90

95

I00

105

I

I

I

h

i

I

EMDGQCKRDLKCCMGMCGKSCVSPVKA

Figure 5. Amino acid sequence analysisof HPC growth-supporting fractions (f) purified from 5637-cell CM. Amino acid sequences were determined by Edman degradation with an on-line Applied Biosystems477A/120A sequencer. Since cysteine residues cannot be detected unless modified, their presence was deducted .from the absence of any detectable amino acid signal at such position. ALP sequence data are availablefrom EMBL/GenBank/DDBJ under accession numbers P03973 and P07757. 1309

Goselink et al.


Figure 4. SDS-PAGE of ILP-HPLC fractions containing biologically active (Fig. 3) 15-kD protein from 5637-cell CM. Of each fraction, 20 p~l was loaded on a polyacrylamide gel under reducing conditions and proteins were silver stained. Molecular mass markers (Bin-Rad, Ikichmond, CA) were phosphorylase b (Mr 92,500), BSA (Mr 66,200), ovalbumin (M, 45,000), carbonic anhydrase (Mr 31,000), soybean trypsin inhibitor (M, 21,500), lysozyme (Mr 14,400), and the low M, marker aprotinin (Pierce Chemical Company, Rockford, 1L) (Mr 6,500).

proteinase inhibitor ALP. ALP consists o f a total o f 107 residues. W h e n purified natural protein was digested with formic acid, sequence analysis o f the internal amino acids 50 to 76 showed again identity to ALP, confirming its c o m plete h o m o l o g y to ALP (Fig. 5). Similarly, the second peak o f activity from gel filtration (fractions 68 and 69) was purified by H P L C , yielding again a predominant 15-kD p r o tein (fraction 28), which was also identified as ALP by NH2-tenmnal sequence analysis (45 residues) (Fig. 5). As shown in Fig. 6, purified rALP could also support the growth o f C F U - G M and B F U - E similar to 5637 C M or the purified 15-kD protein. Polyclonal a n t i - h u A L P antibodies could completely neutralize the biologic activity o f 5637 C M (10% vol/vol), H P L C fraction-32 (0.1% v o l / vol), and rALP (10 ng/ml). H o w e v e r , the support o f AB serum (1% vol/vol) on H P C growth could not be neutralized by polyclonal a n t i - h u A L P antibodies (Fig. 6). l 6 0 % o f the myeloid colony growth as compared to the control culture in the presence o f 10% AB serum. In addition, w e could completely neutralize the biologic activity from 5 6 3 7 - C M and purified natural ALP with a polyclonal IgG a n t i huALP, illustrating that ALP was the factor essential for the in vitro H P C growth from m o n o c y t e - and T l y m p h o c y t e depleted bone marrow cells in the presence o f highly purified h u m a n albumin. ALP is a 107-amino acid cationic protein consisting o f two domains o f similar architecture but with different inhibitory activities (18, 19). T h e NH~terminal domain A L P - D - 1 that is assumed to bind trypsin could not support growth o f H P C in the absence o f serum. The C O O H - t e m f i n a l domain ALP-D-2, containing a strong inhibitory activity against chymotrypsin, neutrophil elastase, and trypsin, restored the colony growth, indicating that the proteinase inhibitory activity o f ALP was indeed responsible for the support o f the H P C growth in vitro. ALP, also k n o w n as secretory leukocyte proteinase inhibitor, is present in a wide variety o f mucous secretions and produced by epithelial-like cells (20). This local p r o duction o f ALP with very high affinity for leukocyte elastase serves as a potential regulatory feedback mechanism to prevent epithelial damage caused by proteolytic proteinases (21). Regulatory effects o f ALP in hematopoiesis have not been described previously. Serum contains various serine proteinase inhibitors, o f which 0r PI and o~2-macroglobulin are present at relatively high concentrations o f 1.3 m g / m l and 1.7 m g / m l , respectively (22); ALP has been found to circulate at a concentration o f 5(i) n g / m l (21). W h e n serum was replaced by c~-1 PI, the roam plasma p r o -



ies. Before initiation of the colony assayAB serum (AB, 1"/,vol/vol), 5637 CM (CM, 10% vol/vol), purified natural ALP (f-32, 0.1% vol/vol), rALP (10 ng/mnl) were incubated at 37~ during 2 h in the presence of control medium or 40 p~gof polyclonal lgG anti-huALP antibodies, sufficient to neutralize 400 ng ofrALP. Colony growth was expressedas percentage of growth of control culture, obtained in the presence of 10% (vol/vol) AB serum. The number of CFU-GM and BFU-E in the control culture was 27 and 28 per 103 cells plated, respectively. In the presence of antihuALP antibodies, colony growth supported by 5637 CM, fraction f-32 or rALP, but not by 1% (vol/vol) AB serum, was completely neutralized.

200


--A-IMDM

g rALP

I CM --l-ABs - 0 - a-1 PI 140

1

2

1

4

0

0

~ -

1

12

16

10 8O

60 40 O

20

12.

B u

2

CFU-GM

0 0

1 2

4

6

8

12

16

160,

0

1 2

'201_.,/%. 1

4

0

6

8

0

~

I

0

0

~

,

6o

:

0'

0

1 2

4

6

8

12

16

CDa4-/CD34+ ratio

0

0

1 2

4

6

1-

8

A

12

16

CD34-/CD34+ ratio

Figure 8. Inhibition of the suppressive effect of CD34- accessory cells on HPC growth of purified CD34 + cells in the absence of serum by proteinase inhibitors. In the presence of HGF, the CD34 + cells were cultured with increasing numbers of CD34- cells, up to a ratio 1:16 supplemented with IMDM, rALP (100 ng/ml), CM (10% vol/vol), AB serum (10% vol/vol), or ct-1 Pl (2 l~g/ml). The growth of CFU-GM and BFU-E is expressed as percentage of growth of control culture, obtained in the presence of 10% (vol/vol) AB serum in the absence of CD34- cells. The results of two experiments (iLl and/3) are shown.

teinase inhibitor of neutrophil elastase (22), the growth support was comparable to rALP. These results support the evidence that the proteinase inhibitory activity is responsible for the observed effects. The high concentrations of other serine proteinase inhibitors may explain why antihuALP antibo :lies did not abolish the stimulatory effect of serum. Since the bone marrow accessory cells responsible for the suppression of growth of CD34-purified H P C were depleted of monocytes and T lymphocytes, the myeloid cells appeared to be the mediator cells for this effect. Probably, these myeloid cells secrete proteolytic proteinases like neutrophil elastase that are capable of degrading cytokines, growth factor receptors, or other proteins essential for the proliferation of H P C (23-25). ALP or other proteinase in-

hibitors present in serum apparently neutralized these proteolytic enzymes. The mechanisms by which ALP supported the proliferation of H P C in vitro may be at the cell membrane level by protecting from degradation cellular receptors or proteins in the culture medium that are essential for H P C growth. In conclusion, we demonstrated that monocyte- and T lymphocyte-depleted accessory bone marrow cells suppressed the in vitro growth o f C D 3 4 + H P C in the absence of serum. The serine proteinase inhibitor ALP was capable of restoring the proliferation of H P C in vitro in the presence of multiple HGF, highly purified albumin, insulin, and cholesterol. These results show that proteinase inhibitors are important for optimal in vitro growth of H P C and may be essential for in vitro expansion of human hematopoietic stem cells in serum-free medium.

The authors thank Yvonne Bauman-Souverein for her help in preparing the manuscript and Willy Put, Jean-Pierre Lenaerts, lken~ Conings, Paul Proost, Maarten van de Keur, and Arie van de Marel for technical assistance. 1311

Goselink et al.


120


Anja Wuyts is a research assistant of the Belgian National Fund for Scientific Research. This study was supported by a grant from the J.A. Cohen Institute for Radiopathology and Radiation Protection and by grants from the National Fund for Scientific Research of Belgium. Address correspondence to Dr. J.H.F. Falkenburg, Department of Hematology, Leiden University Hospital, Building 1:C2-R-140, P.O. Box 9600, 2300 RC, Leiden, The Netherlands. Received for publication 19June 1995 and in revised form 13June 1996.

References

1312

15.

16.

t7.

18.

19.

20.

21.

22.

23.

24.

25.

bridoma growth factor identical to T-cell-derived B-cell stinmlatory factor-2 (interleukin-6). Eur. J. Biodlem. 168: 543-550. Van Damme, J., G. Opdenakker, RJ. Simpson, M.R. Rubira, S. Cayphas, A. Vink, A. Billiau, and A. Van Snick. 1987. ldentifcation of the human 26-kD protein, interferon [32 (IFN-[32), as a B cell hybridoma/plasmacytoma growth factor induced by interleukin-1 and tumor necrosis factor. J. Exp. Med. 165:914-919. Van Damme, J., C. Uyttenhove, F. Houssiau, W. Put, P. Proost, and J. Van Snick. 1992. Human growth factor for murine interleukin (IL)-9 responsive T cell lines: co-induction with IL-6 in fibroblasts and identification as LIF/ HILDA. Eur..J. hnmunol. 22:2801-2808. Proost, P., A. Wuyts, R. Conings, J.-P. Lenaerts, A. BiUiau, G. Opdenakker, and J. Van Damme. 1993. Human and bovine granulocyte chemotactic protein-2: complete amino acid sequence and functional characterization as chemokines. Biodlemistry. 32:10170-10177. Thompson, R.C., and K. Ohlsson. 1986. Isolation, properties, and complete amino acid sequence of human secretory leukocyte protease inhibitor, a potent inhibitor of leukocyte elastase. Pro& Natl. Acad. Sci. USA. 83:6692-6696. Heinzel, R., H. Appelhans, G. Gassen, U. Seem611er, W. Machleidt, H. Fritz, and G. Steffens. 1986. Molecular cloning and expression of cDNA for human antileukoprotease from cervix uterus. Eur. d. Biochem. 160:61-67. Franken, C., C.J.L.M. Meijer, J.H. Dijkman. 1989. Tissue distribution of antileukoprotease and lysozyme in humans, d. Histochem. Cytod~em. 37:493-498. Kramps, J.A., L.N.A. Willems, R. De Water, C. Franken, and J.H. Dijkman. 1991. Bronchial secreting cells and antileukoprotease. In Endothelial and Mucus Secreting Cells. E. Pozzi, editor. Massano, Milano, Italy. 235-245. Bj6rk, P., L. Axelsson, M. Bergenfeldt, and K. Ohlsson. 1988. Influence of plasma protease inhibitors and the secretory leucocyte protease inhibitor on leucocyte elastaseinduced consumption of selected plasma proteins in vitro in man. Scand..J. Clin. Lab. hwest. 48:205-211. Laouar, A., J. Wietzerbin, and B. Bauvois. 1993. Divergent regulation of cell surface protease expression in HL-60 cells differentiated into macrophages with granulocyte macrophage colony stimulating factor or neutrophils with retinoic acid. Int. Immunol. 5:965-973. Deane, D., L. Inglis, and D. Haig. 1995. The 175 antigen expressed on myeloid and erythroid cells during differentiation is associated with serine protease activity. Blood. 85:1215i219. Shipp, M.A., and A.T. Look. 1993. Hematopoietic differentiation antigens that are membrane-associated enzymes: cutting is the key! Blood. 82:1052-1070.



1. Sieff, C.A., S.C. Ekern, D.G. Nathan, and J.W. Anderson. 1989. Combinations of recombinant colony-stimulating factors are required for optimal hematopoietic differentiation in serum-deprived culture. Blood. 73:688-693. 2. Barnes, D., and G. Sato. 1980. Methods for growth of cultured cells in serum-free medium. Anal. Biochem. 102:255-270. 3. Correa, P.N., and A.A. Axelrad. 1991. Production of erythropoietic bursts by progenitor cells from adult human peripheral blood in an improved serum-free medium: role of insulin-like growth factor 1. Blood. 78:2823-2833. 4. Schwartz, G.N., W.R. Hudgins, andJ.F. Perdue. 1993. Glycosylated insulin-like growth factor II promoted expansion of granulocyte-macrophage colony-forming cells in serumdeprived liquid cultures of human peripheral blood cells. Exp. Hematol. 21:1447-1454. 5. Lansdorp, P.M., and W. Dragowska. 1992. Long-term erythropoiesis from constant numbers of CD34 + cells in serumfree cultures initiated with highly purified progenitor cells from human bone marrow.J. Exp. Med. 175:1501-1509. 6. Mayani, H., W. Dragowska, and P.M. Lansdorp. 1993. Characterization of functionally distinct subpopulations of CD34 + cord blood cells in serum-free long-term cultures supplemented with hematopoietic cytokines. Blood. 82:26642672. 7. Krystal, G., V. Lain, W. Dragowska, C. Takahashi, J. Appel, A. Gontier, A. Jenkins, H. Lain, L. Quon, and P. Lansdorp. 1994. Transforming growth factor [31 is an inducer of erythroid differentiation.J. Exp. Med. 180:851-860. 8. Cohn, E.J. 1947. Separation of blood into fractions of therapeutic value. Ann. Intern. Med. 26:341-352. 9. Chen, R.F. 1967. Removal of fatty acids from serum albumin by charcoal treatment. J. Biol. Chem. 242:173-181. 10. Goselink, H.M., D.E. Williams, W.E. Fibbe, H.W. Wessels, G.C. Beverstock, R. Willemze, andJ.H.F. Falkenburg. 1992. Effect of mast cell growth factor (c-kit-ligand) on clonogenic leukemic precursor cells. Blood. 80:750-757. 11. Kramps, J.A., C. Franken, and J.H. Dijkman. 1984. ELISA for quantitative measurement of low-molecular-weight bronchial protease inhibitor in human sputum. Am. Rev. Respir. Dis. 129:959-963. 12. Ruck, A., E. Jakobson, S. Bj6rkman, and S. Paulie. 1994. Adaptation of human bladder carcinoma cell lines to serumfree growth. Evidence for autocrine growth stinmlation. Anticancer Res. 14:55-60. 13. Kaashoek, J.G.J., R. Mout, J.H.F. Falkenburg, R. Wiltemze, W.E. Fibbe, andJ.E. Landegent. 199l. Cytokine production by the bladder carcinoma cell line 5637: rapid analysis of m R N A expression levels using a cDNA-PCR procedure. Lymphokine Cytokine Res. 10:231-235. 14. Van Damme, J., S. Cayphas, J. Van Snick, R. Conings, W. Put. J.-P. Lenaerts, R.J. Simpson, and A. Billiau. 1987. Purification and characterization of human fibroblast-derived hy-