review Haematologica 1995; 80:367-387

CD34-POSITIVE CELLS: BIOLOGY AND CLINICAL RELEVANCE Carmelo Carlo Stella, Mario Cazzola,* Paolo De Fabritiis,° Armando De Vincentiis,# Alessandro Massimo Gianni,@ Francesco Lanza,^ Francesco Lauria,** Roberto M. Lemoli,°° Corrado Tarella,## Paola Zanon,@@ Sante Tura°° Cattedra di Ematologia, Università di Parma, Parma; *Dipartimento di Medicina Interna e Terapia Medica, Sezione di Medicina Interna e Oncologia Medica, Università di Pavia e IRCCS Policlinico S. Matteo, Pavia; °Dipartimento di Biopatologia Umana, Sezione di Ematologia, Università “La Sapienza”, Roma; #Dompé Biotec SpA, Milano; @Centro Trapianto Midollo Osseo “Cristina Gandini”, Divisione di Oncologia Medica C, Istituto Nazionale Tumori, Milano; ^Istituto di Ematologia e Fisiopatologia dell’Emostasi, Università degli Studi di Ferrara, Ferrara; **Istituto di Scienze Mediche, Università degli Studi di Milano, Milano; °°Istituto di Ematologia “Lorenzo e Ariosto Seragnoli”, Università degli Studi di Bologna, Bologna; ##Cattedra di Ematologia, Università di Torino, Torino; @@Amgen S.p.A., Milano, Italy

The CD34-positive cell: definition and Ta ken toget h e r, these findings ref l ect the het- morphology erogeneous proliferative status and syn- Cellular expression of the CD34 antigen iden- thesis of this cell population. Conversely, earlier tifies a morph o l ogi c a lly and immu n o l ogi c a lly hematopoietic progenitors, identified as CD34+ heterogeneous cell population that is function- C D 3 3 – H L A- D R –, seem to be more hom o ge- a l ly ch a racteri zed by the in vi t ro c a p a b i l i t y to neous in size (small lymphocyte-like cells) and gen era t e cl onal aggrega tes derived from early l a ck cytoplasmic gra nules and prom i n e n t and late progenitors and the in vivo capacity to nucleoli. Again, the morphology of this cellular reconstitute the myelo-lymphopoietic system in pop u l a t i o n appe a r s to ref l ect the functi o n a l a supralethally irradiated host.1-3 ch a r acteri s tics of these cells (e.g. low pro tei n Im mu n o h i s toc h emical studies have dem on- synthesis, very low pro l i fera tive activi t y wi t h s tra ted that the CD34 anti gen is stage but not predominantly G0 phase). l i n e a g e specific. In fact, indepen den t ly of the Several mon ocl onal anti bod ies (MY10, 12.8, d i f f eren ti a tive lineage, it is ex pre s s ed on ly by B1-3C5, 115.2, ICH3, TUK3, etc.) ra i s e d on togen eti c a l ly early cell s . 4 For ye a rs a major against the leukemic cell lines KG1 or KG1a and ob s t acle to the morph o l ogical iden ti fic a ti on of an anti - en do t h eli al cell anti body (QBEND10) p ut a tive hem a top oi etic has been the assigned to the CD34 cluster have been shown d i f f i c u l t y in sep a ra ting them from their direct to identify a transmembrane glycoproteic anti- progeny. The use of CD34 and other su i t a bl e gen of 105-120 kD ex pre s s ed on 1-3% of nor- cell surface markers (i.e. CD33, CD38, HLA-DR mal cells, 0.01-0.1% peri ph era l a n ti gens) in flu ore s cen ce - activa ted cell - s orti n g bl o od cells and 0.1-0.4% cord bl o od cell s . 5 techniques or other cell separation methods has Different antibodies recognize distinct epitopes allowed considerable progress in this field. of the same antigen. CD34 antigen expression is Positively selected, lineage committed CD34+ associated with concomitant expression of sev- cells and more immatu r e, lineage nega t ive eral other markers that can be classified as lin- C D 3 4 + C D 3 3 – H L A- D R- cells are shown in eage non-specific markers (Thy1, CD38, HLA- F i g u re 1 and Figure 2, re s pectively. On May - DR, CD45RA, CD71) and lineage spec i f i c Grünwald-Giemsa stained preparations, CD34+ m a r kers, including T- l ym ph oid (T d T, CD10, cells are medium sized cells having large nuclei, CD7, CD5, CD2), B-lym p h oid (T d T, CD10, eccen t ri c a l ly su r ro u n ded by narrow rims of CD19), myeloid (CD33, CD13) and megakary- deep blue cytoplasm occ a s i o n a l ly con t a i n i n g oc y tic (CD61, CD41, CD42b) markers . 5 Th e cytoplasmic granules. Some normal CD34+ cell ex p re s s i on of lineage non - s p ecific markers nu cl e i show one or more pale blue nu cl eo l i . allows the heterogeneous CD34+ population to

Correspondence: Prof. Sante Tura, Istituto di Ematologia L. e A. Seràgnoli, Policlinico S. Orsola, via Massarenti 9, 40138 Bologna, Italy. Acknowledgments: the preparation of this manuscript was supported by grants from Dompé Biotec SpA, Milano, and Amgen S.p.A., Milano, Italy. Received October 18, 1994; accepted May 2, 1995. 368 C. Carlo Stella et al.

Figure 1. May-Grünwald-Giemsa (MGG) staining of cytospin preparation of Figure 2. MGG staining of cytospin preparation of enriched CD34+ CD33– enriched CD34+ cells, highly purified by avidin-biotin immunoaffinity. HLA-DR– cells. The CD34+ cell fraction was further depleted of CD33+ HLA- DR+ cells by immunomagnetic separation. be divi ded into two disti n c t su bpop u l a t i o n s tively small pop u l a t i on of stem cells, loc a ted ch a r acteri z ed, re s p ectively, by low or high mainly in the bone marrow, that can (i) under- expression of Thy1, CD38, HLA-DR, CD45RA, go sel f - ren ewal to produ ce more stem cells or CD71. These two cell su bpop u l a ti ons con t a i n (ii) differentiate to produce progeny which are e a r ly and late hem a topoi etic progen i tor cell s , progre s s ively unable to sel f - ren ew, irrevers i bly respectively.6-8 com m i t t ed to one or another of the va r i o u s In ad d i t i on to conven t i onal immu n o l ogi c a l h em a topoi e tic lineages, and able to gen e ra te m a r kers cl a s s i fied on the basis of their assign a - cl ones of up to 105 l i n e a ge - re s tri cted cells that ti o n to specific clu s ters of differen t i a t i o n , mature into specialized cells.16-18 CD34 cells ex press receptors for a nu m ber of The dec i s i o n of a stem cell to ei t h e r sel f - growth factors. Two disti n ct families of rel a ted ren e w or differen t i a te and the sel ecti o n of a receptors have been iden t i f i e d: (i) tyro s i n e specific differentiation lineage by a multipotent kinase receptors, including the stem cell factor progen i tor du ring com m i tm e nt are intri n s i c receptor (SCF-R, CD117) and the mac roph a ge properties of stem cell progenitor cells and are co l ony - s ti mu l a ting factor receptor (M-CSF- R , reg u l a ted by stoch a s tic mech a n i s m s . 1 9 Su r viva l CD115); (ii) hem a top oi etic receptors not con- and amplification of hematopoietic progenitors taining a tyrosine kinase domain, su ch as the are controlled by a number of regulatory mole- gra nu l o c y te - m ac r oph a ge co l ony - s ti mu l a t i n g cules (hem a topoi etic growth factors) interact- f actor receptor (GM-CSF-R, CDw116).5 , 9 ing according to com p l ex mod a l i ties (syner- The iden ti fic a ti on of new markers sel ectively gism, rec r u i t m ent, antagon i s m ) . 1 9 A furt h e r ex pre s s e d on pri m i t ive lym ph o h em a topoi eti c l e vel of hem a topoi e tic con t rol is exerted by cells (CD34+ C D 3 8 –) repre s ents a sti m u l a ti n g nu clear tra n s c r i p ti on factors that activa te lin- re s e a rch field. In this con text, stem cell tyro s i n e e a g e - s pecific genes reg u l a t ing growth factor kinase receptors (STK), su ch as STK-1, a hu m a n responsiveness and/or the proliferative capacity h o m o l ogue of the mu r ine Flk-2/Flt-3, are of of hematopoietic cells.20 p a r ticular rel e va n ce . 1 0 - 1 2 The ligands for these Detection of the most primitive hematopoiet- receptors might repre s ent new factors able to ic cell types is now po s s i ble due to the tech- s el ectively con trol stem cell sel f - ren ewal, pro l i f- nique of long-term bone marrow culture. In the era ti on and differen ti a ti on . 1 3 - 1 5 case of human bone marrow, a 5- to 8-wee k time peri o d bet ween initi a t ing cultu res and assessing clonogenic progenitor numbers allows Clonogenic and biologic activity qu a n t i f i c a ti o n of a very pri m i tive cell in the The structural and functional integrity of the starting population, the so-called long-term cul- h em a topoi etic sys tem is maintained by a rel a- ture-initiating cell (LTC-IC).21 Committed prog- CD34-positive cells 369 enitors of the various hematopoietic cell classes lym ph oid progen i tors wh i ch are not assaya bl e can be qu a n ti t a ted by a nu m ber of short - term with current in vitro systems. Non-proliferating c u l t u re cl o n o genic assays . 1 9 CD34 anti g en C D 3 4 + cells might repre s ent a su bpop u l a ti on ex p re s s i on assoc i a ted with low CD38 and that is not re s pon s ive to conven ti onal myel oi d C D 4 5 R A ex p re s s i o n and va r i a b le HLA- D R h em a topoi etic growth factors. The non - pro l i f- expression is a typical feature of LTC-IC, CFU- era t ing CD34+ su b s et might requ i re the pre s- Blast, CFU-T, CFU-B. In con trast, CD34 anti- ence of co-factors, such as the ligand of STK-1 gen expression associated with CD38 and HLA- or the hepatocyte growth factor, able to activate DR ex pre s s i on is a typical fe a tu re of mu l ti po- s tem cell - s p ecific genes whose ex pre s s i on is a tent (CFU-GEMM) and lineage - re s t ri c ted prerequ i s i t e for acqu i r ing re s p on s iveness to (CFU-GM, CFU-G, CFU-M, BFU-E, CFU-E, conventional growth factors.14,22,23 B F U - Meg, CFU-Meg) hem a topoi etic progen i- In let h a l ly irrad i a ted non - human pri m a te s , tor cells5 (Figure 3). Recently reported data have both auto l ogous and all ogen eic CD34+ cells have s h own that low or absent ex pre s s i o n of the been shown to have the capac i ty to recon s ti tut e Thy1 or SCF receptor can be efficiently used to the myel o - lym ph opoi etic sys tem, thus su gge s t- en r i c h pri m i t ive hem a topoi e tic progen i tors ing that the stem cell re s pon s i ble for hem a topoi- f rom the heterogen e ous CD34+ cell pop u l a - etic recon s ti tuti on is CD34+.2 4 , 2 5 It has also been ti on . 7 , 9 Al t h o u g h the CD34 anti gen is vi rtu a lly s h own that human CD34+ H LA- D R – cells tra n s- expressed by all progenitor cells, the percentage p l a n ted in utero in the fetus of sheep initi a te of CD34+ cells with assaya b le in vi t ro cl o n o- and sustain a ch i m eric hem a topoiesis produ c i n g genic activity ranges from 10 to 30%. The prob- human progen i tor cells of all differen ti a tive lin- lem of non-clonogenic CD34+ cells is still open e a g e s . 2 6 Auto l o gous CD34+ cells en r i ch e d by and not adequately explained by the presence of avi d i n - bi o tin co lumns have been shown to be a b le to recon s t i tute myel o - lym ph o poiesis in p a t i e nts receiving high - d ose ch e m orad i o - t h era py.2 7 The re sults of studies using CD34+ bone marrow cells in the all ogen eic set ting in p a ti ents receiving both rel a ted as well as unrel a t- ed all ogen eic marrow transplants wi l l soon be ava i l a bl e . 2 7 In ad d i ti on, trials are planned that wi ll use all ogen eic peri ph eral bl ood CD34+ cell s ei t h er alone or with marrow.2 7

Characterization and function of the CD34 cell surface molecule The CD34 cell su rf ace molecule has been bi o- ch em i c a lly ch a racteri zed and both the hu m a n c D NA and gene have been cl on ed and sequ en c- ed in the last few ye a rs . 2 8 , 2 9 CD34 is a one-pass type I tra n s m em brane gly- copro tein with a molecular wei ght of 105-120 kDs in ei t h er the redu ced or unredu ced form2 8 ( F i g u re 4). CD34 pro tein is not hom o l ogo us to a ny other known pro tein. The minimum size of the CD34 pro tein is 354-amino acids and con- tains nine sites for N-glyco s yl a t i on and a severa l for O-glyco s yl a ti on that are essen t ial con s ti- Figure 3. Cellular organization of the hematopoietic system. tu ents of the three ep i topes of the molecule; this 370 C. Carlo Stella et al.

shown that the human CD34 gene is located on 1, and recent studies with in situ hybri d i z a t i on have assign ed its loc a l i z a ti on to band 1q32, in close prox i m i t y to other gen e s that en code growth factors or functi on mole- cules su c h as CD1, CD45, TGF2, laminin, LAM/GMP, etc.28 Seven CD34 mon ocl onal anti bodies (Mo Ab s ) were clu s tered at the 4th Work s h op on Leu ko- c y te Di f feren ti a ti on An ti gens (Vi enna, 1988),3 0 , 3 1 and another 15 Mo Abs were veri fied as recog- nizing the CD34 molecule du r ing the 5th In tern a ti onal Work s h op on Leu koc y te Di f feren- Lipid bilayer ti a ti on An ti gens (Bo s ton, 1983), the most direct evi den ce being re activi ty with cells tra n s fec ted with CD34 cDNA and binding to CD34 GPI anchor pro tei n . 3 2 , 3 3 The ep i tope spec i fic i ty of the CD34 a n t i b odies was cl a s s i f i e d into three disti n c t N-glycosylation site groups according to the sen s i tiv i ty of the ep i- O-glycosylation site topes to en z y m a tic cl e ava ge (wh i ch was per- form ed using neu raminidase, chym opapain and Figure 4. Schematic representation of the CD34 cell surface molecule. glycopro tease from Pa s teu r ella haem olyti c a ) , re activi t y with fibroblasts and high en do t h el i a l Fig. 4 venules, and cross bl ocking ex peri m ents (Ta bl e m o l ecu le is also ri ch in sialic acid. Its bi och em i- 1 ) . 3 4 , 3 5 We know, in fact, that glycopro tease from cal com po s i ti on su ggests a mu c i n - l i ke stru ctu re Pa s teu rella haem olyti c a s pec i f i c a lly cl e ave s on ly and in some re s p ects re s e m bles leu c o s i a l i n pro t eins containing sialyl a t ed O-linked gly- (CD43). Sequ en ce com p a r i s ons bet ween hu m a n c a n s . 3 4 Ba s e d on these data, it can be furt h e r and mouse CD34 show a very low level of iden- po s tu l a ted that class III ep i topes are more prox i - ti t y in the glyco s yl a ted regi on, 70% iden ti ty in mal to the ex tracellular side of the cell mem- the gl o bular domain, and 92% in the tra n s - brane than class I and class III ep i topes. m em brane and cytoplasmic regi ons. Fu rt h e rm ore, for most CD34 Mo Abs (wi t h Using a KG1 cell line libra r y, it has been few excepti ons) cross bl ocking ex peri m ents are

Epitope class Clones CD34 reactivity

paraffin frozen western section section blotting

I. Sensitive to neuraminidase (from Vibrio 14G3, BI3C5, My10,* 12.8,* positive negative positive cholerae), chymopapain, glycoprotease (from ICH3,* Pasteurella haemolytica) Immu-133, Immu-409

II. Resistant to neuraminidase. Glycoprotease, 43A1, MD34.3, positive positive positive and chymopapain sensitive MD34.1, MD34.2, QBend10, 4A1, 9044, 9049

Table 1. Epitope specificity of CD34 III. Resistant to neuraminidase, chymopapain CD34 9F2,HPCA2, negative positive negative MoAbs as assessed by their differential and glycoprotease 581, 553.563, Tuk 3, 115.2 sensitivity.

*Incomplete digestion by neuroaminidase. CD34-positive cells 371 in agreem ent with the cl a s s i f i c a ti on based on dem on s t ra ted that CD34 prob a bly acts as an en z y m a t ic cl e ava ge of the CD34 pro tein. In ad h e s ive ligand for L- s el ectin. It has been fur- o t h er words, using a cocktail of CD34 Mo Ab s , t h er po s tu l a ted that the CD34 molecule co u l d CD34 re activi ty is bl ocked on ly in the case that p l a y a pro t ective role against pro t eo l yti c Mo Abs bel on ging to the same CD34 ep i tope are en z ym e - m ed i a ted damage due to its high nu m- s i mu l t a n eo u s ly em p l oyed. On the con tra ry, the ber of O-glyco s yl a t i on sites. The cytop l a s m i c com bi n ed use of Mo Abs recognizing class II and domain contains two sites for pro tein kinase C III or class I and II or class III ep i topes does not ph o s ph oryl a ti on and one for tyrosine ph o s ph o- a f f ect cell re activi t y. Moreover, CD34 Mo Ab s ryl a ti on . 2 8 defining class III ep i top es are unable to re act Given the discordant reactivity of these mole- with CD34 glycopro t ein in We s t ern bl o t s cules, the choice of the CD34 MoAb to employ because this ep i tope is sen s i tive to den a tu r a ti on. may be important when analyzing cell positivi- The pattern of expression of CD34 antibodies ty for the CD34 molecule in both leukemic and exhibited by CD34+ acute leukemias is partially normal samples, and may be responsible for the in accord a n ce with that derived from ep i top e d i f feren ces reported by va rious aut h ors in the mapping based on the differential sensitivity of literature concerning the prognostic role played CD34 to enzymatic treatment. However, about by this anti gen in ac ute myel oid leu kem i a s . 3 7 one third of CD34 Mo Abs do not seem to The type of CD34 Mo Ab used to enu m era te bel ong to any of these su b g roups and for this progen i tor cells is prob a b ly rel e vant in the peculiar pattern of expression are referred to as peri ph era l bl ood stem cell autograft set ting as atypical CD34 reagents. The widest variation in well, since both early and late engraftment fol- CD34 Mo Ab re activi ty has been dem on s tra ted l o wing tra n s p l a n t a t i on are, to some ex ten t , in ac u te myel o id leu kemia (AML) samples, related to the number of hemopoietic stem cells allowing us to postulate the occurrence of aber- collected at the time of or bone marrow rant anti g ens or of disti n ct ep i topes in su b - h a r vests, and to the degree of progen i tor cell groups of leu kemias. Al tern a tively, it can be maturation related to the expression of lineage hypo t h e s i z ed that the ex p re s s i on of differen t markers such as HLA-DR, CD71, CD38, CD33, a n ti bodies could ref l ect the degree of matu ra- and myeloperoxidase. tion of leukemic cells. The presence of new, dis- ti n c t non overl a pping ep i top es could be pro- posed on the basis of the data published so far Techniques for CD34+ cell separation in the litera tu re. As far as the ex pre s s i on of A nu m ber of different tech n i ques have been CD34 in normal and leu kemic cells is con- proposed for separating CD34+ cells. The com- cerned, it has been calculated by flow cytometry m on aim is to produ ce a cell pop u l a ti on wi t h that the nu m b er of molecular equ iva l ents of optimal purity and viability by means of a low s o l u b le flu o roch r ome (MESF) ex pre s s e d by cost, rapid and simple sep a r a t i on tech n i qu e . l eu kemic and normal progen i tors ra n ges from The first separation techniques exploited para- 18,200 to 322,000 and from 8,000 to 124,000, m e ters su ch as size and cell den s i t y and were respectively. represented by Ficoll-Hypaque and Percoll den- Recent data co ll ected by Lanza et al.3 6 s eem to s i t y grad i ents. In the last two dec ades, the i n d i c a te that class I-type Mo Abs are more sen s i- devel o pm ent of mon o cl o nal anti b odies has tive to freezing procedu res than class II and III a l l owed a more specific and careful cell sel ec- Mo Abs, since the ep i top e is not iden t i fia ble fo l- tion by identifying surface antigens used as tar- l owing a frozen / t h awed met h odo l ogy. gets for cell separation (Table 2). The functi on of CD34 su r f ace glycopro tein in h em a topoi etic stem and progen i tor cells is sti ll FACS (Fluorescence Activated Cell Sorter) the obj ect of deb a te. In of recent fin d i n gs , F l ow cytom etry is able to phys i c a lly sep a ra te it would seem to play a rel evant role in modu- d i f ferent pop u l a ti ons after incubati on of cell s l a ting cell ad h e s i on . 3 6 Fu rt h erm ore, it has been with flu o roch r om e - con ju g a t ed mon o cl o n a l 372 C. Carlo Stella et al. antibodies. This cell sorting technique can yield Table 2. Recovery of CD34-positive cells after different separation tech- a high ly puri f i e d (> 98%) cell pop u l a ti o n. In niques. ad d i t i on, the use of el e ctronic ga t es all o ws Enrichment Recovery Large-scale s e l ecti on and recovery of several su b pop u l a - separation tions according to antigenic expression and dif- (% CD34+ (% CD34+ in the recovered of the original ferent characteristics such as size and cytoplas- population) sample) mic gra nu l a ri t y. This tech n i que has been very + Negative depletion useful for studying CD34 sub-populations but by immunomagnetic 20-60 30-60 no cannot be employed to select large numbers of beads cells due to its com p l ex i ty and low recovery. Positive selection by The recent development of high-speed cell sort- immunomagnetic 60-80 30-60 yes beads ing, however, might allow clinical utilization of this technique. Fluorescence activated cell sorter (FACS) > 95 30 - 50 time consuming Panning Panning 50 - 80 30 - 60 yes

An t i-CD34 mon ocl onal anti bodies bound to Ceprate SC® 50 - 80 40 - 70 yes one of the su rf aces of cell cultu re flasks were recently used to select CD34+ cells. When a cell suspension is introduced into the flask, the pos- i t ive pop u l a ti on is bl ocked on the plastic su r- ti on and dep l eti on of spec i fic cell pop u l a ti on s . 4 0 f ace, while CD34 nega tive cells remain in su s- The instru m ent inclu des a set of non - reu s a bl e pension and can be easily eliminated. Adherent produ cts including bi o ti nyl a ted anti CD34 anti- cells should contain the CD34+ population with body, plastic bags, filters and a co l umn of a viability >90%.38 avi d i n - bi o tin be ads. An autom a ted vers i on con- tro ll ed by a com p uter wh i ch guara n tees repro- Immunomagnetic systems du c i bi l i t y of opera ti o n and redu ces risks of Im mu n om a g n etic be ads are uniform, su per- opera tor errors has been devel oped for cl i n i c a l p a ra m a g n etic, po lys t yrene be ads with affinity use. Its capac i ty has recen t ly been incre a s ed so purified anti-mouse Ig covalently bound to the that a single co lumn can process more than su rf ace. Th ey are equ a lly su i ted for nega t ive 5 0 31 0 1 0 bone marrow or peri ph eral bl ood cell s and positive cell separation; the rosetted target in 1-2 hours and sustain bone marrow en gra f t- cell can easily be isolated by applying a magnet m ent in pati ents su bm i t ted to autogra f t . 4 1 on the outer wall of the test tube for 1-2 min- utes. Im m u n o m a g n e tic be a ds co u p l e d wi t h CD34 mon o cl onal anti b odies can be uti l i z ed CD34-positive subpopulations: phenotypic for positive selection of CD34+ cells to obtain a and functional analysis pop u l a t i o n with > 80% vi a ble CD34+ cell s . 3 9 The normal CD34+ cell population likely con- Si m i l a r ly, immu n o m a g n e tic be a ds can be tains progen i t ors of all human lym p h o - em p l oyed for nega tive dep l e ti on with mon o- h e m a topoi e tic lineages, including stem cell s cl onal anti bodies binding lineage - s pec i fic anti- c a p a ble of hem a topoi etic recon s ti tuti on after gens. bone marrow tra n s p l a n t a ti on . 1 Levels of CD34 ex pre s s i on decline with differen ti a t i on; con s e- High-affinity chromatography based on biotin- qu e n t l y, the earliest cl on o genic cells (CFU- avidin interaction blast, LTC-IC, etc.) express the highest levels of This met h od is based on an immu n oad s orp- CD34, while the most differen ti a ted (CFU-G, ti on tech n i que that relies on the high affinity CFU-M, CFU-E, CFU-Meg) ex press on ly low i n teracti on bet ween the pro tein avidin and the l e vels of CD34 (Figure 5). The CD34 anti g en vitamin bi o tin. Avi d i n - bi o tin immu n o ad s orp- has been used to identify, enumerate and isolate ti on has been em p l oyed for both po s i tive sel ec- cells from different lym ph o - h em a topoi etic lin- CD34-positive cells 373 eages, as well as develop in vitro tests for indi- animals tre a ted, all showed com p l ete marrow rect evaluation of cells with different functional engraftment after reinfusion of Ia-positive cells; and clonogenic capacity.42 on ly one dog died from infecti on . 4 5 Si m i l a rly , m a r row cells from 5 pri m a tes (baboons) were Pre-clinical studies tre a ted in vi tro with a bi o ti nyl a ted anti - C D 3 4 Several animal-human systems have been cre- antibody and then passed through a column of a ted to uti l i ze ch i m eras for the stu dy of lym- avidin; after autograft, all the animals showed ph o - h em a topoiesis in vivo. The first ex p eri- marrow engraftment followed by hematological m e nts dem on s t ra t ed the fe a s i b i l i t y of tra n s - recon s t i tuti on com p a ra b le to that of con t ro l planting human fetal stem cells to sheep fetuses; animals.46 Furthermore, the demonstration that the po s tnatal pre s en ce of human cells in the a ll ogen eic CD34+ cells can recon s ti tute the he- s h e ep was doc u m e n t ed at several points in m a t opoi e tic sys t em in let h a l ly irrad i a t ed ti m e . 4 3 Fu rt h erm ore, some early CD34+ su b - b a b oons con f i rm e d that this cell pop u l a t i on populations were able to repopulate sheep bone includes pluripotent stem cells.25 m a r row; animals were tra n s p l a n ted in utero with CD34+/ D R – cells and the pre s e n ce of a Lymphoid precursor cells ch i m eric pop u l a ti on with the functi onal ch a r- The CD34+ cell com p a rtm ent contains all the acteri s t ics of hem o poi e tic progen i t ors was cells ex pressing terminal deox y nu cl eo ti dyl tra n s- dem on s tra ted in the marrow and peri p h era l ferase (T d T ), wh i ch is an intra nu clear en z ym e blood in a percentage of cases.44 Berenson et al.45 ex pre s s ed in early lym ph oid cells under goi n g also showed how hem a topoi etic progen i tors i m m u n o gl o bulin or T- c ell receptor gen e (positive for the Ia antigen and subsequently for re a rra n gem ent. Flow cytom etry has shown that the CD34 antigen) could reconstitute the mar- the great majori ty of TdT+ cells in the marrow row of let h a l ly irrad i a ted dogs. Of the seven coex press CD34, CD19 and CD10 (B-cell pre-

Figure 5. Functional differentiation and antigenic expression of hematopoietic cells. 374 C. Carlo Stella et al. c u rs ors), as well as T- cell differen ti a ti on anti- Multilineage progenitors and stem cells gens su ch as CD7, CD5 and CD2. A small pro- CFU-GEMM contain prec u rs or cl o n ogen i c porti on of CD34+/ T d T + cells coex press CD10, cells of both myel o id and eryt h roid lineage s wh i ch might repre s ent a com m on lym ph o i d and ex p ress CD34, HLA-DR, CD38, CD117 progen i tor for both B and T lineage s .4 7 Recen t ly, and CD45RA. Th ey also ex press low amounts Mi ll er et al. reported that CD34+ cells may also of CD33, but not CD13. In a hypothetical dif- gen era te NK cells in vi tro.4 8 ferentiation scheme involving pluripotent stem cells, the lympho-hemopoietic compartment is Granulocyte-macrophage precursors the next cell type and can be identified in vitro Ma rrow eryt h roid progen i tors lack spec i f i c with CFU-Blast and LTC-IC. m a r kers and therefore are difficult to iden ti f y. The lack of ex pre s s i o n of CD38 is the most G lycoph orin A- d i rected mon ocl onal anti bod i e s important characteristic of these early progeny, recogn i ze all hem ogl obi n i zed cells, but this mol- wh i ch repre s e nt 1% of CD34 and less than ecule is ex pre s s ed in on ly a small proporti on of 0.01% of mononuclear cells. The lack of CD38 C D 3 4 + cells and is absent in cl on ogenic cell s . 4 9 a l l ows sep a r a ti o n of com m i t t ed progen i t ors Hi gh levels of CD45 are pre s ent on BFU-E, ( C D 3 4 +/ C D 3 8 +) from earl i er com p a r tm en t s but this anti gen is lost by the CFU-E stage ; 5 0 ( C D 3 4 +/ C D 3 8 –) by a single marker com bi n a - h owever, the CD45RO isoform is well repre- tion.54 s e n ted on earl i er eryt h roid progen i tors, wh i l e Fu r t h erm ore, the earliest CD34+ cells coex- the CD45RA isoform is pre s ent on com m i t ted press low levels of CD45RO6 and are nega tive myel oid progen i tors. The ex pre s s i on of CD71 for staining with the fluorescent dye rhodamine () is currently considered to 123. The role of HLA-DR in defining earlier cell be the specific antigen for the CD34+ erythroid types is sti ll con troversial; a series of evi den ce pop u l a ti on. CD71 is pre s ent at high levels on i n d i c a tes that the stem cell com p a rtm e nt has erythroid progenitor cells and at very low levels the CD34 +/ C D 3 8 –/ H L A- D R + ph e n o t ype . 8 , 5 4 in all the other progenitor cells.51 Expression of Recent works, however, have not found HLA- CD71 increases from stem cells to BFU-E, then DR in the earliest cells.55 These data were con- declines during erythroid maturation. In addi- firmed by the possibility that HLA-DR expres- ti o n, marrow CD34+ eryt h r oid cells migh t sion may discriminate the Ph-positive leukemic ex press IL-3, GM-CSF (CD116) and eryt h ro- com p a rtm ent (HLA- D R +) from normal re s i d- poi etin receptors, based on the acti on of these ual cells (HLA- D R – ) in ch r onic myel o i d growth factors on CFU-erythroid cells.52 leukemia.56 Myel oid prec u rs o r cl o n ogenic cells (CFU- GM, CFU-G, CFU-M, CFU-MK) coex pre s s CD34, HLA-DR, CD117 (c - k i t ), CD45RA , Adhesion molecules and cytokine receptors CD33 and CD13; CD15 is present at low levels A nu m ber of molecules within the integri n on CFU-G, while CFU-M spec i f i c a l ly ex pre s s family have been shown to mediate interactions CD115 (M-CSF receptor). CFU-MK are the bet ween early CD34-po s i t ive cells and bon e only CD34+ cells which express the platelet gly- m a r row stromal cells. These inclu d e VLA- copro teins iden ti f i ed by the CD61, CD42 and 4 / VCAM-1, VLA- 5 / f i b ron e ctin, LFA-1 or CD41 mon ocl onal anti bod i e s . 5 Den d ri tic cell s ICAM, and several others. Each adhesion mole- also originate from bone marrow, but the con- cule appe a rs to med i a te a spec i fic cell interac- ditions that direct their growth and differentia- ti on. Hem a topoi e tic growth factors may be ti o n are sti l l poorly ch a r acteri zed. GM-CSF active in a solu b le form or in a mem b ra n e - s t i mu l a tes the growth of den d r i tic cells from a n c h ored form; ad h e s i o n molecules may be mouse peri p h e ral bl o od; however, it was crucial for allowing anchored growth factors to recently reported that CD34+ cells may give rise bind the target cell. Table 3 lists the main adhe- to den d ri ti c / L a n gh erans cells after sti mu l a ti on s i o n molecule and growth factor receptors with GM-CSF and tumor necrosis factor-a.53 expressed on CD34-positive cells. CD34-positive cells 375

Table 3. Adhesion molecule and growth factor receptors expressed on of ep i topes on cryopre s erved secti ons, class III CD34-positive cells. ep i top es were not recogn i zed by specific anti - CD34 Mo Abs on paraffin em bed ded secti on s . Antigen name CD Ligand Levels of CD34 expression, highest in immature h e m a topoi etic prec u r s or cells, dec rease pro- GlyCAM-L/ none adhesion molecule gressively with cell maturation. ICAM1,2,3 CD11a/CD18 adhesion molecule Rega rding hem a to l ogic malignancies, CD34 is H-CAM CD44 adhesion molecule ex pre s s e d in a large percen t a ge of ac ute leu ke- VLA-4 CD49d/CD29 adhesion molecule m i a s . 6 1 The flu ore s cen ce inten s i t y of CD34 VLA-5 CD49d/CD29 adhesion molecule ex pre s s i on is va ri a ble and high er in ac ute lym- FGF-R none growth factor ph o bl a s tic (ALL) than in ac ute myel obl a s t i c l e u kemia (AML). In these latter pati ents, the M-CSF CD115 growth factor CD34 anti gen is found on 40-60% of leu kem i c GM-CSF-R CD116 growth factor blasts and is most frequ en t ly assoc i a ted wi t h SCF (c-) CD117 growth factor M0, M1 and M4 FAB cyto t ype, secon d a r y Interferon g-R CD119 growth factor l eu kemias, karyo typic abn orm a l i ties invo lvi n g IL 7-R CD127 growth factor ch rom o s ome 5 or 7, P170 ex pre s s i on and poor progn o s i s . 6 1 - 6 4 Thus, CD34 ex pre s s i on may be con s i der ed the most pred i ctive nega tive factor in AML pati ents stri ct ly correl a ted with inten s i t y CD34 expression on normal and neopl a s tic cell s of ex pre s s i on . 6 5 , 6 6 In ALL, CD34 is ex pre s s ed in It has been known that CD34 mon ocl o n a l 70% of pati ents, parti c u l a rly in those with a B- a n t i bodies bind spec i f i c a l ly to vascular en do- l i n e a g e ph en o t ype. In these pati e nts, unlike t h e l ium ever since a 110 kd pro tein ex tracted AML cases, the clinical rel e va n c e of CD34 from freshly isolated umbilical vessel endotheli- ex pre s s i on is con t roversial; however, accord i n g um was iden ti f i e d with CD34 anti b odies in to the fin d i n gs of a Ped i a tric Onco l ogy Gro u p,6 7 Western blots and in Northern blot analysis.57,58 its ex pre s s i on in B-lineage cases was assoc i a ted CD34 molecules have a striking ultrastructural with hyperd i p l oi dy, lower frequ ency of initi a l l oc a l i z a t i on on en do t h eli al cells: they are con- cen tral nervous sys tem (CNS) leu kemia and a centrated primarily on the luminal side, in par- f avora ble prognosis. In T- cell ALL cases, on the ticular on membrane processes, many of which o t h er hand, CD34 ex pre s s i on showed a po s i tive i n t erd i g i t a te bet w een ad j acent en d o t h el i a l correl a ti on with initial CNS leu kemia and CD10 cell s . 5 7 , 5 8 Si n ce this regi on is an important site n e ga tivi t y, but not with any pre s en ting favor- for leu koc y te ad h e s i o n and tra n s en d o t h el i a l a bl e - risk ch a racteri s ti c s .6 7 , 6 8 tra f fic, in con trast to previous ex peri en ce s , 3 3 i t Lastly, CD34 and HLA-DR expression may be has been hypo t h e s i z ed that CD34 may be very useful in discri m i n a ting bet ween the very antagonistic or inhibitory to the adhesive func- few benign primitive hematopoietic progenitors ti o ns of vascular en do t h el ium. This was su p - and their malignant co u n terp a r ts in pati en t s ported by the dem on s tra ti on that CD34 gen e with chronic myeloid leukemia (CML). In fact, ex pre s s i on at the mRNA level is rec i proc a lly it has been demonstrated that normal progeni- down - r eg u l a t ed wh e n ad h e s i o n molec u l e s tor cells are CD34+ and HLA-DR–, while malig- ICAM-1 and ELAM-1 are up-regulated by IL-1 nant progen i tor cells, wh i ch ex h i b it Ph and du ring inflammatory skin lesions assoc i a ted bc r / a bl re a rra n gem ent, ex press HLA-DR anti- with leu koc y te infiltra ti o n . 3 3 , 5 9 Fu r t h erm o re , gens.56 ad d i ti ona l studies con du cted on both para f f i n em bed ded and cryopre s erved secti ons dem on- s t ra ted that fibroblasts also re act with anti - Positive and negative regulators of hematopoi- CD34 Mo Ab s . 6 0 However, it should be noted etic progenitor cells that while CD34 MoAbs reacted with all classes The is defined by its 376 C. Carlo Stella et al. ex ten s i ve sel f - ren ewal capac i t y, mu l t i l i n e a g e (CSFs) (e.g. gra nu l o c y te - C S F, gra nu l o c y te - d i f f eren ti a t i o n po t en t ial and capac i t y for of m a c roph a g e-CSF), and interl e ukins (ILs). l on g - term recon s t i tuti on of normal marrow Depending on their target cells and differen t f u n c ti o n in let h a l ly irrad i a t ed animals. 6 8 proliferative potential, cytokines can be divided Tra n s p l a n t a ti on of retrovi ra lly marked mu r i n e i n to three categori e s : 7 1 1) late - acti n g , lineage - s tem cells has shown that on ly a few mu l ti l i n - specific factors; 2) intermediate-acting, lineage- eage progenitor cells induce the repopulation of nonspecific factors; 3) early-acting growth fac- en g ra f ted hem a t opoi e tic ti s s u e , 6 9 su g ge s t i n g tors inducing the recruitment of dormant early that hematopoiesis may be supported by a suc- progenitors in the cell cycle (Figure 6).71 cession of short-lived clones. Moreover, experi- The majori ty of late - acting factors su pport the mental evidence indicates that the processes of pro l i fera t i on and matu ra ti on of lineage - com- s el f - r en e wal, differen t i a ti on and sel ecti on of m i t ted progen i tors and the functi onal proper- lineage potentials are intrinsic properties of the ties of term i n a lly differen ti a ted cells. Eryt h ro- s tem cells and occur according to a stoch a s ti c poi etin (Epo) is the phys i o l ogic reg u l a tor of ery- ( r a n dom) model . 7 0 In the ste ady state, most of t h ropoi e s i s 7 3 and throm b opoi etin has the same 7 4 the stem cells are qu i e s cent (G0 phase) and role in throm b opoi e s i s , while M-CSF and IL- 5 begin active cycling ra n dom ly.7 1 Convers e ly, a re con s i dered specific for mac roph a g es and su r vival and pro l i f era ti o n of hem a topoi e ti c eo s i n ophils, re s pectively. G-CSF exerts its activi- progen i tors are reg u l a ted by cytokines, wh i ch ty on com m i t t ed neutrophil prec u r s o rs , also act in preventing apoptosis.72 According to a l t h o u g h it has been shown to be a syner gi s ti c this model, the induction of differentiation by a f a ctor for pri m i tive hem a topoi e tic cell s . 7 5 cytokine may be considered as the consequence In term ed i a te - acti n g, lineage - n on s pecific fac- of the pro l i f era ti o n of a specific pop u l a ti o n tors inclu d e IL-3, IL-4 and GM-CSF. Th ei r s t i m u l a t ed by that factor. The broad term activi t y is mainly directed tow a rd progen i tor cytokines includes growth factors such as fibrob- cells in the intermediate stages of hematopoietic last growth factor, col o ny sti m u l a t ing facto r s development. However, they interact with later-

Figure 6. Cytokines exert their activity at different levels of the hematopoietic differentiation pathway. Each progenitor cell is concurrently affected by multiple regulators. CD34-positive cells 377 acting growth factors for the produ cti on of k i n e s . 2 3 Com bi n a ti ons of two or more growt h m o re matu r e cell s , 7 6 as well as with the f a ctors8 1 can sti mu l a te hem a topoi etic cells ei t h er c y tokines capable of tri g gering stem cell by amplifying the progeny cell produ c ti on of cycling. On their own, they act as survival fac- s i n gle prec u rs ors (syner gy) or by inducing ad d i- tors for stem cells and appear to stimulate early ti o nal cl on ogenic cells to pro l i fera te (rec ru i t - h em a top oi etic prec u rs ors on ly after their ex i t m ent). Examples of these two types of en h a n ce- 77 + from G0. m ent are given in Figures 7 and 8, wh ere CD34 Several cytokines have recen t ly been iden ti- C D 3 3 – D R – cells are simu l t a n eo u s ly sti mu l a ted fied for their capacity to stimulate the prolifera- by two or three growth factors. The molec u l a r ti on of the earliest hem a topoi etic cells. Ma pp- basis reg u l a ting the com p l ex interp l ay bet ween ing studies of normal blast cell co l ony form a- c y tokines is sti ll largely unknown. However, on e ti on from single progen i tors have shown that propo s ed mechanism for growth factor syner- IL-6, G-CSF, IL-11, stem cell factor (SCF), IL-12 gism is the indu cti on of CSF receptors on early and l eu kemia inhibi to ry facto r (LIF) recruit dor- h em a topoi etic progen i tor cell s . 8 2 This appe a rs to mant stem cells into the cell cycle that are then be a coord i n a te cascade tra n s activa ti on via up- a ble to re s pond to ad d i ti onal growth factors . 7 7 m o du l a ti on of growth factor receptors that lead s Wh ereas the perm i s s ive factors retain limited to pro l i fera ti on and differen t i a ti o n of hu m a n pro l i f era t ive po ten t ial by them s e lves, they m a r row cell s . 8 3 Convers e ly, the stru c tu r a l strongly enhance the stimulatory activity of IL- h om o l ogy bet ween some growth factors , 8 4 t h e 3, GM-CSF, G-CSF and EPO on CD34+ a n d pre s en ce of shared receptor su bunits on the cell more immature CD34+ lineage-progenitors.78 In m em bra n e 8 5 and com m on sign a l - t ra n s du cti on addition to positive interaction with intermedi- a te-and late - acting growth factors, SCF syner- gi s ti c a l ly or ad d i tively augm ents the co l ony - forming abi l i t y of other early - acting growt h f a ctors, su c h as IL-11, IL-12 and IL-6, on myel oid and bi l i n e a ge (i.e. lym p h omyel oi d ) pri m i tive cell s . 7 9 Recen t ly, the ligands for the STK-1 or FLT3 receptor, and the hep a toc y te growth factor were shown to be able to stimu- l a te very pri m i t ive hem a topoi etic progen i tor cells. However, their bi o l ogical activi t y is sti ll A under investigation. The main sources of both positive and nega- tive reg u l a tor y pro teins are acce s s ory myel oi d cells and the stromal com pon ent of the bon e marrow. In general, microenvironment cells do not con s t i tutively produ c e cytokines, ra t h er tra n s c r i pti on and/or tra n s l a ti on processes are rapidly induced by cytokines such as IL-1, TNF. The extracellular matrix also participates in the regulation of hematopoiesis by binding growth f a ctors and pre s en ting them in a bi o l ogi c a l ly active form to bone marrow progenitor cells.80 B Most data su g gest that stem cells ex press low l e vels of growth factor receptors and requ i re Figure 7. Human CD34+ CD33- DR- cells were stimulated by IL-9 (A) or IL-9 and SCF (B) in the presence of EPO. The addition of SCF induced the mu l tiple pro l i fer a tive sti muli to en ter into the growth of large multicentered BFU-E colonies containing > 10,000 cells. cell cycle, while com m i t ted progen i tor cells can The different size of the colonies indicates amplification of stem cell prog- be ef f ectively sti m u l a ted by indivi dual cyto- eny (synergy). 378 C. Carlo Stella et al.

human BM prec u r s o rs and its activi t y on h em a topoiesis is on ly inhibi tory,8 9 a l t h o u gh the s y n er gi s tic growth factors IL-11 and IL-9 seem to be able to oppose the nega tive reg u l a ti on of TG F-b3 on human CD34+ cell s . 9 1 Several po ten- tial modes of acti on of the TG F-b f a m i ly have been su gge s ted, including down - m odu l a ti on of c y tokine receptors , 9 2 i n teracti on with the under- ph o s p h oryl a t ed form of the reti n o bl a s t om a 9 3 gene produ ct in late G1 ph a s e , and altera ti on of Figure 8. The addition of SCF to IL-11, in the presence of EPO, results in a gene ex pre s s i on . 9 4 E a rly studies have shown that much higher number of colony-forming cells derived from CD34+ CD33– b – TG F- 1 and 3 exert their activi ty on norm a l DR progenitors (right), as compared to the IL-11 and EPO combination and leu kemic cells by len g t h eni ng or arre s t i n g (left) (recruitment). 9 5 the G1 phase of the cell cycl e , and this ef fect is f u n cti onal to pro tect normal CD34 po s i tive cell s pro tei n s 8 6 provi de a po ten tial ex p l a n a ti on for f rom the tox i c i t y of alkyl a ting agents in vi tro.9 6 t h e ir functi onal similari ties and the app a ren t More recent inve s ti ga ti ons have dem on s tra ted redundancy of their activi ty. that TG F-b reg u l a tes the re s pon s iveness of mice The growth of hem a topoi etic progen i tor cell s h em a topoi etic cells to SCF, wh i ch is known to is also reg u l a ted by solu ble nega tive reg u l a tors be the main syner gi s tic factor for both mu ri n e su ch as mac roph a ge inflammatory pro tei n - 1a and human stem cells, thro u gh a dec rease in c - ( M I P- 1 a), tu m or nec rosis factor-a (T N F-a) , kit m R NA stabi l i t y that leads to dec re a s ed cell - i n t erferons (IFNs), pro s t a glandins and tra n s - su rf ace ex pre s s i on . 9 7 forming growth factor-b (TG F-b). The TG F-b Similarly to TGF-b, TNF-a has been reported f a m i ly of pro teins inclu des at least five isoform s to have both inhibitory and stimulatory effects (TG F-b1-5) wh i ch are en coded by differen t on hem a t opoi e tic progen i tor cells. TNF-a gen e s 8 7 and produ ced by stromal cells, platel et s po ten t i a ted the IL-3 and GM-CSF- m ed i a ted and bone cells. Moreover, a su b s et of very pri m- growth of human CD34+ cells in short-term liq- i tive mu rine hem a topoi etic cells has been shown uid cultu re assay. However, it inhibi t ed the to sec rete active TG F-b1 by an autoc rine mech- growth prom o t ing activi t y of G-CSF.9 8 Two a n i s m . 8 8 TG F-b1 and 2 isoforms are bi m od a l TNF receptors with molecular wei g hts of 55 reg u l a tors of mu r ine and human hem a topoi eti c and 75 kd, respectively, have recently been iden- progen i tor cells, and their activi ty is based upon ti f i e d . 9 9 Wh e reas the p55 receptor med i a t e s the differen ti a ti on state of the target cells and T N F-a ef fects on com m i t ted progen i tor cell s , the pre s en ce of growth factors . 8 9 In the hu m a n the p75 receptor is invo lved in signaling the s ys tem , for instance, CFU-GEMM derived from inhibition of murine primitive cells. p u r i f i ed CD34+ C D 3 3 – cells are inhibi t ed by Fu r t h e rm o re, it was recen t l y shown that TG F-b1, wh ereas more com m i t ted progen i tors T N F -a is capable of inhibi t ing the mu l t i - su ch as CFU-G or CFU-GM are not affected. In growth factor (GM-CSF, IL-3, G-CSF, SCF, IL- ad d i t i on, high pro l i f era t ive po ten t i a l - C F C 1 ) - d epen d ent growth of human HPP- C F C ( H P P-CFC) re s p on s ive to a com b i n a t i on of derived from CD34+ cells thro u gh interacti on CSFs are markedly inhibi ted by TG F-b1, wh i l e with both p55 and p75 receptors, while the p55 m o re matu re CFU-GM are actu a lly en h a n ced receptor exclu s ively med i a tes the bi f u n cti on a l wh en GM-CSF is used as co l ony forming fac- activity of TNF-a on more mature BM precur- tor.7 8 TG F-b1 and 2-indu ced myel o su ppre s s i on sors responsive to single cytokines.100 is parti a lly co u n teracted by early - acting growt h M I P- 1a is a pepti de of 69-amino acids with a f a ctors su c h as G-CSF, IL-6 and fibrobl a s t m o l ecu lar wei ght of 7.8 kd produ ced by activa t- growth factor.9 0 On the other hand, TG F-b3 has ed mac roph a ges, T- cells and fibrobl a s t s . 1 0 1 It been shown to be a more po tent su ppre s s or of bel on gs to a large family of put a tive cyto k i n e s CD34-positive cells 379 that inclu des MIP- 1b, MIP-2 and IL-8 (ch e m o- al bl ood samples taken under ste ady state con d i- kines). Bi o l ogic ch a racteri z a ti on has shown that ti ons. Large qu a n ti ties of CD34+ cells can be co l- M I P- 1 a en h a n ces the M-CSF- and GM-CSF- l ected with massive marrow harvests, su ch as for depen dent growth of CFU-GM, while it inhibi t s tra n s p l a n t a ti on purposes. Nevert h eless, marrow the co l ony - forming abi l i t y of hem a topoi eti c C D 3 4 + cells are som ewhat elu s ive due to thei r prec u rs ors pre s ent in a cell pop u l a ti on en ri ch ed s c a t t ering among the predominant CD34+ for CD34+ + D R + cells sti mu l a ted with eryt h ro- h em a topoi etic pop u l a ti on . 1 0 5 Recen t ly devel oped poi etin, IL-3 and GM-CSF.1 0 2 cell sep a r a t i on procedu res all ow co l l ecti on of Ta ken toget h er, these re sults indicate that a h i g h l y en r i c h e d CD34+ cell pop u l a t i o n s . com p l ex interp l ay bet ween po s i tive and nega tive However, this is gen era lly accom p l i s h ed thro u gh reg u l a tor y pro teins determines the pro l i fera ti on a s pec i fic and of ten unaccept a ble cell loss.1 0 6 So far or inhibi ti on of early hem a topoi etic progen i tor the limited nu m b er of immatu re prec u r s o rs cells (Figure 9). In gen e ral, the activi t y of com m on ly obt a i n ed from both bone marrow i n h i b i t ors of hem o poiesis appe a r s to be and peri ph eral bl ood has been the major ob s t acl e revers i ble, lineage - n on s pec i fic and directed at the to a simple iden ti fic a ti on and analysis of marrow e a rly stages of differen ti a ti on. In ad d i ti on, TG F-b C D 3 4 + cells. In d eed, the growing interest in has shown some degree of differen tial activi t y C D 3 4 + cells is pri m a r i ly the re sult of the devel op- bet ween normal and neop l a s tic lym ph oid cell s . 9 6 m ent of new thera p eutic mod a l i t ies that all ow Thus, nega tive reg u l a tors may be cl i n i c a lly rel e- easy access to large qu a n t i ties of hem opoi e ti c vant to the pro tecti on of the hem a topoi etic stem prec u rs ors thro u gh the peri ph eral bl ood. The key cell com p a rtm ent from the do s e - l i m i ting tox i c i ty role in these innova tive approaches is repre s en ted of neop l a s tic disease thera py.9 6 , 1 0 3 , 1 0 4 by the introdu cti on of hem opoi etic growth fac- tors for clinical use.1 0 7 At pre s ent GM-CSF and G-CSF are the most Coll e ction of CD34+ cell s ex ten s ively em p l oyed and stu d i ed hem opoi eti c Bone marrow and peri p h eral bl ood are the growth factors in the clinical set ti n g. From the on ly sources of immatu re hem opoi etic prec u r- very begi n n i n g, it was ob s erved that GM-CSF or s ors iden ti fied as CD34+ cells. A diagn o s tic mar- G-CSF ad m i n i s t ra ti on was assoc i a ted with an row sample contains on l y a few CD34+ cell s , i n c rease in circ u l a t ing hem opoi e tic progen i - while even fewer of them are pre s ent in peri ph er- tors . 1 0 8 , 1 0 9

Differentiation

Self renewal

Negative Positive Survival

IFN a, b, g TGF-b1-2 SCF IL-11 TGF-b3 TNF-a IL-11 GM-CSF MIP-1a IL-12 IL-6 G-CSF Figure 9. Interplay between positive and IL-1 negative regulatory acting on IL-9(?) hematopoietic stem cells. 380 C. Carlo Stella et al.

L a ter on, it was dem on s tra ted that this ph e- ch em o t h era py that indu ces profound leu koc y- n om en on could be ex ten s ively and reprodu c i bly topenia seems to be crucial for optimal mobi- a m p l i fied by com bining growth factor ad m i n i s- l i z a t i o n; for instance, cycl o ph o s p h a m i d e at tra ti on with high - dose ch em o t h era py.1 1 0 - 1 1 2 doses lower than 7 g/sqm produ ces a redu ced In deed hem opoi etic progen i tors are massively, m obilizing sti mu lu s . 1 1 1 , 1 2 2 Several other ch em o- t h o u g h tra n s i en t ly, mobi l i zed into peri ph era l t h e ra py sch edules have also been su cce s s f u lly bl ood du ring hem opoi etic recovery fo ll owi n g em p l oyed for mobi l i z a ti on purposes. The pri n- h i g h - dose ch em o t h era py given with growth fac- cipal ch e m o t h era py pro tocols reported to be tor su pport. Su ch an abu n d a n ce of immatu re h i g h ly ef fective in inducing CD34+ cell mobi l i z a- h em opoi e tic cells makes them easily recogn i z- ti on are su m m a ri zed in Ta ble 4.110, 111, 115, 117, 122-127 a ble by cell sorting tech n i ques that em p l oy anti - As stre s s ed earl i er, hem opoi etic growth factors CD34 Mo Ab s . 1 1 3 Un der optimal con d i ti ons, the p l ay a key role in mobi l i z a ti on. This has been proporti on of CD34+ cells may re a ch va lues as cl e a rly doc u m en ted with cycl oph o s ph a m i de. A h i g h as 20-30% of the total leu koc yte count. In m e dian peak va lue of 75 CD34+ cells/µL has s t e a dy state con d i t i ons CD34+ cells do not been recorded fo ll owing high - dose cycl oph o s- exceed 4% of the total marrow pop u l a t i on , ph a m i de alone, wh ereas 420 and 500/µL are the while they are undetect a ble by cell sorting tech- m edian va lue s recorded wh en GM-CSF or G- n i ques in the peri ph eral bl o od. Ch em o t h era py C S F, re s p ectively, are ad d ed to cycl o ph o s - and growth factors do not merely indu ce a rel a- ph a m i d e . 1 1 2 , 1 1 3 , 1 2 8 , 1 2 9 Ex ten s ive growth factor- tive increase of immatu re hem opoi etic cell s ; i n du ced mobi l i z a ti on is furt h er su b s t a n ti a ted by t h eir absolute nu m ber is amplified several ti m e s the po s s i bi l i t y of co l l ecting su f f i c i ent CD34+ over basal con d i t i ons. This all ows co ll ecti on of cells using growth factor alon e . 1 3 0 , 1 3 1 In this set- su f fic i ent amounts of hem opoi etic progen i tors ting the most promising ex peri en ces have been for autogra f ting purposes by means of a few l eu k a ph eresis procedu re s .1 1 0 , 1 1 3 , 1 1 4 3 4 Va lues of 10-20 1 0 CFU-GM/kg repre s en t Table 4. Main chemotherapy protocols employed in hematopoietic progeni- the minimal required dose for marrow engraft- tor mobilization. ment with peripheral blood progenitors.115-117 In f a ct, mu c h high e r qu a n t i t ies of circ u l a t i n g Authors Protocol Drug(s) Dosage progen i tors can be co ll ected using appropri a te characteristics employed mobilization procedures. Under optimal condi- ti ons, circ u l a ting CD34+ cell peak va lues may Gianni et al.110 single agent cyclophosphamide high range between 150 and 700/µL on days of maxi- 123 6 Tarella et al. single agent etoposide high mal mobi l i z a t i o n. As a rule, at least 831 0 Gianni et al.124 C D 3 4 + cells/kg or more can thus be co ll ected with 1 to 3 leu k a ph eresis procedu re s . 1 1 4 Th e s e Kotasek et al.121 single agent cyclophosphamide intermediate Tarella et al.122 hu ge qu a n ti ties, approx i m a tely corre s pon d i n g to 503104 CFU-GM/kg, must be considered the Dreyfus et al.125 single agent cytarabine high i deal threshold dose of peri ph eral bl ood prog- Schimazaki et al.126 multiple drugs etoposide high enitors for autografting purposes. Indeed values cytarabine of 831 0 6 C D 3 4 + cells/kg or more guara n tee a Kawano et al.115 multiple drugs daunorubicin intermediate rapid and du ra ble hem opoi etic recovery wh en cyclophosphamide c i r c u l a t ing progen i t ors are used as the sole etoposide s o u rce for marrow recon s t i tuti on fo l l owi n g 127 118-121 Pettengell et al. multiple drugs doxorubicin intermediate submyeloablative treatment. cyclophosphamide Thus far, massive CD34+ cell mobi l i z a ti on has etoposide been most com m on ly ob s erved wh en growt h Hass et al.117 multiple drugs cytarabine high f a ctor is ad m i n i s t ered fo l l owing high - d o s e mitoxantrone c y cl oph o s ph a m i de, given at 7 g/sqm. In deed CD34-positive cells 381 produ ced with G-CSF. The re s ults reported ters . 1 1 4 , 1 4 0 , 1 4 3 , 1 4 4 The harve s ted cells are re su s pen ded i n d i c a te new opportu n i ties for the uti l i z a ti on of in freezing med ium and then cryopre s erved for m obi l i zed progen i tors in all ogen eic tra n s p l a n t a - su b s e qu e nt tra n s p l a n t a t i on. One to 3 leu k a - ti on procedu re s . 1 3 2 , 1 3 3 Com bi n ed ch em o t h era py ph ereses repe a ted on con s ec utive days usu a lly and growth factors remain the most ef f i c i en t provi de large amounts of progen i tor cells capa- m obi l i z a ti on procedu re at this time. However, ble of rapid en gra f tm ent after su bmyel oa bl a tive on going studies are directed at improving mobi- tre a tm en t s . l i z a ti on ef fic i ency with growth factors alone. For Factors and procedu res favoring CD34+ cell example, G-CSF at doses high e r than 5 m o bi l i z a t i o n have been well establ i s h e d . µ g / k g / d ay up to 20 µg/kg/day may have high er However, other con d i ti ons advers ely influ en c i n g m o bi l i z a t i on capac i ty.1 3 4 Fu r t h er improvem en t the mobi l i z a ti on ph en om en on should be con- m i g ht derive from the clinical ava i l a bi l i ty of new s i d ered. A major limitati on is repre s en ted by c y tokines, su ch as IL-3, SCF and others, to be i m p a i red marrow functi on, as can occur in pre- em p l oyed alone or in com bi n a ti on with G- or vi o u s ly tre a ted pati e n t s . 1 1 1 In fact, pati ents at G M - C S F.1 3 5 - 1 3 7 However, the po ten ti a l ly high first relapse fo ll owing a single tre a tm ent pro to- m obilizing activi ty of su ch new cytokine com bi- col maintain an adequ a t e mobi l i z a t i o n n a ti ons must be accom p a n i ed by no or very few c a p ac i ty;1 4 5 h owever, few if any mobi l i zed CD34+ s i de ef fects in order to be con s i dered for a wi de cells can be obt a i n e d from heavi l y tre a ted clinical app l i c a bi l i ty. p a ti ents previ o u s ly ex po s ed to mu l tiple ch em o- Mobilizing pro tocols gen era l ly inclu de daily t h era py co u rses. Mobi l i z a ti on is also profo u n dly del ivery of growth factors, starting 1 to 3 days redu ced and of ten to t a l ly abo l i s h ed in pati en t s a f t er ch em o t h e ra py ad m i n i s tra t i on and con- previ o u s ly ex p o s ed to rad i o t h era py, espec i a lly if ti nuing until harve s ting procedu res are com- it was del ivered to the pelvis or to ex ten ded ver- p l e ted. Growth factor is usu a lly ad m i n i s tered tebral are a s . 1 1 7 , 1 2 3 L a s t ly, marrow inva s i on by for a total of 7-12 con s e c utive days. The most tu m or cells may nega tively affect mobi l i z a ti on conven i ent ro ute of del ivery is su bc ut a n e o u s , c a p ac i ty. This is typ i c a lly reported in myel om a with 1 to 2 doses per day. Progen i tor cell har- p a ti ents in wh om the ex tent of CD34+ cells of ten vests are perform ed du ring hem opoi etic recov- correl a tes with the degree of marrow invo lve- ery, provi ded that progen i tor cell mobi l i z a ti on m ent by tu m oral plasma cell s . 1 2 2 In con clu s i on , is doc u m e n ted. In d eed va r ious para m e ters s e veral new fin d i n gs have dra m a ti c a lly improv- h ave been con s i dered as an indirect indicati on ed the procedu r es for co l l e cti o n of large of progen i t or mobi l i z a t i o n, including an amounts of CD34+ cells. However, optimal mar- i n c r ease in WBC or, altern a t ively, in mon o - row functi on is a prerequ i s i te for ex p l oi ting fully c y tes, basophils or platel e t s . 1 3 8 - 1 4 0 However, the activi ty of all those factors known for thei r C D 3 4 + cell eva lu a ti on remains mandatory for m obi l i z a ti on - i n ducing capac i ty. an acc u r a te defin i ti on of the ex tent of progen i- tor mobi l i z a t i on . 1 4 1 , 1 4 2 C D 3 4 + cells should be mon i tored daily from CD34+ positive cells in blood the early stages of hem o poi e tic recovery. Significant numbers of human hematopoietic Detecti on of circ u l a ting CD34+ cells is not su f fi- stem cells can be found in umbilical cord blood c i ent re a s o n for starting leu k a ph eresis proce- and can be used for all ogen eic bone marrow du res; an adequ a te nu m ber of CD34+ cells (>20- tra n s p l a n t a t i on. Mayani et al.1 4 6 found that 1- 50/µL) and WBC > 1 . 0 31 0 9/L and platel et co u n t 2% of the total nu m ber of cord bl ood - der ived > 3 0 31 0 9/L are requ i red for safe and ef fective l ow - d en s i t y cells ex p ress high levels of the progen i tor cell harve s ti n g .1 1 4 , 1 4 2 Leu k a p h ere s e s CD34 antigen and low or undetectable levels of a re perform ed using con ti nu o u s - flow bl ood cell the antigens CD45RA and CD71. These popu- s ep a ra tors. A com p l ete leu k a ph eresis procedu re lations were highly enriched in clonogenic cells gen era l ly takes 2-3 hours and the total bl ood (34%), in particular in multipotent progenitors vo l ume proce s s e d ra n g es from 6 to 10 li- (12%). By cultu ring these cells at low con cen- 382 C. Carlo Stella et al. tra ti on for 8-10 days in high ly defin ed seru m - which occurs after circulating progenitor auto- free liquid cultures supplemented with various tra n s f u s i o n (CPAT) has not been form a l ly h e m a topoi e tic cytokines, it was po s s i ble to proved, but it is most likely a con s equ en ce of ach i e ve a significant ex p a n s i o n (abo ut 100- the mu c h larger (10- to 100-fold high e r ) fold) of the CD34+ cell population. amount of com m i t ted progen i tors rei n f u s e d These last results were recently confirmed by wh e n a pati e nt is autogra f t ed with bl o od - Traycott et al.1 4 7 in an el egant stu dy showi n g derived (as oppo s ed to marrow - derived) cell s . that umbilical cord bl ood CD34+ cells ra p i dly The most likely hypothesis is that these late exit G0-G1 phases and start to cycle in response progen i tors (po s t - progen i tors) of gra nu l oc y te s to stem cell factor. This fe a tu re would make and platel e ts are capable of giving rise to cord bl ood CD34+ cells more su i t a b le candi- mature progeny within a few days, thus allow- dates than bone marrow cells for ex vivo expan- ing submyeloablated patients to survive the ini- sion. tial aplastic phase. It is clear that in vitro expansion and matura- The fundamental role of com m i t ted progen i- tion of hematopoietic progenitor cells might be tor cells was el ega n t ly proved in mice by Jon e s of particular rel e va n ce for tra n s p l a n t a ti o n of et al.1 5 0 None of the let h a lly irrad i a ted animals cord blood hematopoietic cells; however, infor- tra n s p l a n ted with a pure pop u l a t i on of stem m a ti on abo ut the ef fects of su ch an ex p a n s i on cells free of more matu re progen i tors (CFU- on the cells required for long-term hematopoi- GM, CFU-S) su r vived the initial aplasia. A etic reconstitution is highly desirable. m o re recent paper1 5 1 ch a l l en ged this ‘co nven- tional wi s d o m’ m o del, and maintained that peri ph eral bl ood stem cells per se a re capable of The dual role of peri p h eral bl ood hem a topo i eti c ra p i dly matu ring in vivo. progen i tor cells in oncoh em a tol ogy Ot h er aut h ors, using a very similar approach , C D 3 4 + progen i t or cells circ u l a t ing in the re ach ed an opposing con clu s i on . 1 5 2 In hu m a n s peripheral blood represent an enriched and eas- the most convi n c i n g , albeit indirect, evi den ce ily accessible source of two distinct cell popula- in favor of the role of com m i t ted progen i tors in tions: committed progenitor cells and hemato- accel era ting po s t - transplant recovery was pro- poietic stem cells. vi ded by Robert s on et al. , 1 5 3 who doc u m en ted a Al t h o u g h circ u l a t ing progen i tors are com- s i g n i fic a n t ly pro l on ged hem a topoi etic recovery monly called stem cells, the presence of circulat- for pati ents under going auto l ogous bone mar- ing stem cells has been formally proved only in row tra n s p l a n t a t i on purged ex vivo with anti - mice.148 In humans the presence of hematopoi- CD33 mon ocl o nal anti b odies. This re su l t , etic stem cells in the peripheral blood is almost wh i ch occ u rred after a tre a tm ent that sel ective- certain (see below), but to call reinfusion of cir- ly kills the most matu r e progen i t or cell s culating progenitors ‘transplantation’ of periph- ( ex pressing the CD33 su rf a ce anti gen), repre- eral bl o od stem cells (PBSC or similar s ents convincing indirect proof of the role of acronyms) is hardly appropriate. This terminol- com m i t ted progen i tors in early en gra f tm en t . ogy overl o oks the fact that the trem en d o u s The clinical role of com m i t ted progen i tors in interest in peripheral blood autografting is not reducing the morbi d i ty and mort a l i ty of high - (at least so far) a con s equ en ce of its being a dose thera py has ex p a n ded since hem opoi eti c simple su rroga te for auto l ogous bone marrow growth factors have become cl i n i c a lly ava i l a bl e . tra n s p l a n t a ti on (as the term PBSC would su g- In fact, infusion of rhGM-CSF or rh G - C S F, in gest), but rather derives from the unique prop- p a r ticular after ad m i n i s t ra ti o n of myel o tox i c erty of this procedure to reduce the duration of doses of certain stem cell - s p a ring agents, all ows the severe pancytopenia that fo ll ows su bmye- easy co ll ecti o n of an amount of CFU-GM/kg l o a b l a t ive tre a t m ents from two - t h ree wee k s body wei ght 10 to 100 times high er than that ( w h en bone marrow cells are used) to a few con t a i n ed in a bone marrow harve s t . 1 1 0 days only.110,149 The reason for the rapid recovery As alre ady doc u m en ted by a large nu m ber of CD34-positive cells 383 p a p ers, the use of circ u l a ting progen i tors has References bro u g ht abo ut a dra m a tic ch a n ge in the per- 1 . Civin CI, Strauss LC, Brovall C, Fackler MJ, Schwartz JF, s p ectives of high - dose su bmyel oa bl a tive regi- Shaper JH. Antigenic analysis of hematopoiesis III. A m ens. In fact, these on ce spec i a l i zed, ex pen s ive hematopoietic progenitor cell surface antigen defined by a and high ly toxic tre a tm ents are now well to l er- monoclonal antibody raised against KG-1a cells. J Immunol 1984; 133:157-65. a t ed, easy to ad m i n i s ter, and cl i n i c a l ly usef u l 2 . Sutherland RD, Keating A. The CD34 antigen: structure, ( cost ef fective). As an example, it is worth men- biology and potential clinical applications. J Hematother 1992; 1:115-29. ti oning the initial Milan Ca n cer In s ti tute ex pe- 3 . Andrews RG, Singer JW, Bernstein ID. Precursors of colony- ri en c e in a group of over 50 poor- r isk bre a s t forming cells in humans can be distinguished from colony- c a n cer pati e nts who received high - dose mel- forming cells by expression of the CD33 and CD34 antigens and light scatter properties. J Exp Med 1989; 169:1721-31. phalan plus an optimal amount of circ u l a ti n g 4 . Andrews RG, Singer JW, Bernstein ID. Monoclonal antibody progen i t ors ( 531 0 4 C F U - G M / k g body 12.8 recognizes a 115-kd molecule present on both unipotent wei g ht). These pati ents requ i red a median of and multipotent hematopoietic colony-forming cells and their precursors. Blood 1986; 67:842-52. 10.5, 11 and 12.5 days to score > 0.5, 1.0 and 5 . Civin CI, Gore SD. Antigenic analysis of hematopoiesis: a 2 . 0 31 0 9 /L neutrophils/µL, re s p ectively. review. J Hematother 1993; 2:137-44. 6 . Lansdorp PM, Sutherland HJ, Eaves CJ. Selective expression Moreover, more than half of them did not of CD45 isoforms on functional subpopulations of CD34+ requ i re platel et tra n s f u s i ons, while the rem a i n- hematopoietic cells from human bone marrow. J Exp Med ing ones needed on ly one or two tra n s f u s i on s 1990; 172:363-6. 7 . Baum CM, Weissman IL, Tsukamoto AS, Buckle AM. du ring the first week after autogra f ti n g . Su ch Isolation of a candidate human hematopoietic stem-cell pop- mild to modera te tox i c i ty was never de s c ri bed ulation. Proc Natl Acad Sci USA 1992; 89:2804-8. before the clinical ava i l a bi l i t y of com m i t ted 8 . Huang S, Terstappen LWMM. Lymphoid and myeloid differ- entiation of single human CD34 +, HLA-DR –, CD38– progen i tor cells. hematopoietic stem cells. Blood 1994; 83:1515-26. In con c lu s i on, born as a ‘co m pa s s i o n a te’ su r- 9 . Gunji Y, Nakamura M, Osawa H, et al. Human primitive roga t e of bone marrow autogra f ti n g , tod a y hematopoietic progenitor cells are more enriched in KITlow cells than in KIThigh cells. Blood 1993; 82:3283-9. C PAT is ra p i dly rep l acing ABMT. In fact, tod ay 10. Rosnet O, Marchetto S, de Lapeyriere O, Birnbaum D. Muri- it is the latter that should be con s i dered a ‘co m- ne Flt3, a gene encoding a novel tyrosine kinase receptor of the PDGFR/CSF1R family. Oncogene 1991; 6:1641-50. pa s s i o n a te need ’ procedu re, useful in those few 11. Matthews W, Jordan CT, Wiegand GW, Pardoll D, Lemi- p a ti ents unable to mobi l i ze a su f fic i ent nu m ber schka IR. A receptor tyrosine kinase specific to hematopoietic of circ u l a ting progen i tor cell s . stem and progenitor cell-enriched populations. Cell 1991; 65: 1 1 4 3 - 5 2 . The second, disti n ct role of circ u l a t ing prog- 12. Small D, Levenstein M, Kim E, et al. STK-1, the human en i tors is rel a ted to the pre s en ce of stem cell s , homologue of Flk-2/Flt-3, is selectively expressed in CD34+ i.e. to t i p o t ent prec u r s o rs re s p on s i b le for human bone marrow cells and is involved in the proliferation of early progenitor/stem cells. Proc Natl Acad Sci USA 1994; du r a ble recon s ti tut i on of all lym ph o - h em a to- 9 1 : 4 5 9 - 6 3 . poi e tic lineages fo l l owing marrow abl a tive 13. Lyman SD, James L, van den Bos T, et al. Molecular cloning of a ligand for the flt 3 / flk-2 tyrosine kinase receptor: a prolif- t h era py. Si n ce vi r tu a lly no high - dose tre a tm en t erative factor for primitive hematopoietic cells. Cell 1993; 75: that can be safely ad m i n i s t ered to humans is 1157-67. genu i n ely myel oa b l a tive, formal proof of the 1 4 . Lyman SD, James L, Johnson L, et al. Cloning of the human homologue of the murine flt3 ligand: a growth factor for ex i s ten ce of circ u l a ting stem cells must aw a i t early hematopoietic progenitor cells. Blood 1994; 83:2795- ei t h er stable tra n s du cti on of DNA markers into 8 0 1 . a utogra f ted cells or the use of this cell pop u l a- 1 5 . Hannum C, Culpepper J, Campbell D, et al. Ligand for FLT3/FLK2 receptor tyrosine kinase regulates growth of ti on for all ogra f ti n g. haematopoietic stem cells and is encoded by variant RNAs. These experiments are presently underway in Nature 1994; 368:643-8. s everal labora tor i e s , 1 5 4 , 1 5 5 and prel i m i n a r y data 1 6 . McCulloch EA. Stem cells in normal and leukemic hemo- poiesis (Henry Stratton lecture, 1982). Blood 1983; 62:1-13. do con f i r m the pre s en ce of stem cells in the 1 7 . Metcalf D. The molecular control of cell division, dif- peripheral blood of humans. Their utilization is ferentiation commitment and maturation in haemopoietic cells. Nature 1989; 339:27-30. ex pected to revo luti on i ze fields like all ogen ei c 1 8 . Orlic D, Bodine DM. What defines a pluripotent hematopoi- bone marrow transplantation and somatic gene etic stem cell (PHSC): will the real PHSC stand up! Blood therapy, whenever the target of genetic manipu- 1994; 84:3919-94. 156 1 9 . Ogawa M. Differentiation and proliferation of hematopoietic lations is the hematopoietic cell. stem cells. Blood 1993; 81:2844-53. 384 C. Carlo Stella et al.

20. Tsai FY, Keller G, Kuo FC, et al. An early haematopoietic TB, Rosenberg SA. Separation and growth of human CD4+ defect in mice lacking the transcription factor GATA-2. and CD8+ tumor infiltrating lymphocytes and peripheral Nature 1994; 371:221-6. blood mononuclear cells by direct positive panning on cova- 2 1 . Sutherland HJ, Eaves CJ, Eaves AJ, Dragowskas W, Lansdorp lently attached monoclonal antibody coated-flasks. J Biol PM. Characterization and partial purification of human mar- Resp Modifiers 1990; 9:463-74. row cells capable of initiating long-term hematopoiesis in 3 9 . Civin CI, Strauss LC, Fackler MJ, Trismann TM, Wiley JM, vitro. Blood 1989; 74:1563-70. Loken RL. Positive stem cell selection. Basic science. In: S 2 2 . Kmiecik TE, Keller JE, Rosen E, van de Woude GF. Hepato- Gross, A Gee, DA Worthington White, eds. Bone Marrow cyte growth factor is a synergistic factor for the growth of Purging and Processing. New York: Wiley-Liss, 1990:387- hematopoietic progenitor cells. Blood 1992; 80:2454-7. 4 0 2 . 2 3 . Metcalf D. Hematopoietic regulators: redundancy or subtle- 4 0 . Berenson RJ, Bensinger WI, Kalamasz D. Elimination of ty? Blood 1993; 82:3515-23. Daudi lymphoblasts from human bone marrow using avidin- 2 4 . Berenson RJ, Andrews RG, Bensinger WI, et al. CD34+ m a r- biotin immunoadsorption. Blood 1986; 67:509-15. row cells engraft lethally irradiated baboons. J Clin Invest 4 1 . Berenson RJ, Bensinger WI, Hill RS, et al. Engraftment after 1988; 81:951-9. infusion of CD34+ marrow cells in patients with breast cancer 2 5 . Andrews RG, Bryant EM, Bartelmez SH, et al. CD34 marrow or neuroblastoma. Blood 1991; 77:1717-22. cells, devoid of T and B lymphocytes, reconstitute stable lym- 4 2 . Carlo-Stella C, Mangoni L, Piovani G, Garau D, Almici C, phopoiesis and myelopoiesis in lethally irradiated allogeneic Rizzoli V. Identification of Philadelphia-negative granulo- baboons. Blood 1992; 80:1693-9. cyte-macrophage colony-forming units generated by stroma- 2 6 . Srour EF, Zanjani ED, Cornetta K, et al. Persistence of adherent cells from chronic myelogenous leukemia patients. human multilineage, self-renewing lymphohematopoietic Blood 1994; 83:1373-80. stem cells in chimeric sheep. Blood 1993; 82:3333-42. 4 3 . Zanjani ED, Pallavicini MG, Ascensao JL, et al. Engraftment 2 7 . Berenson R. Transplantation of hematopoietic stem cells. J and long-term expression of human fetal hemopoietic stem Hematother 1993; 2:347-9. cells in sheep following transplantation in utero. J Clin Invest 2 8 . Greaves MF, Brown J, Molgaard HV, et al. Molecular features 1992; 89:1178-88. of CD34: a hemopoietic progenitor cell-associated molecule. 4 4 . Srour EF, Zanjani ED, Brandt JE, et al. Sustained human Leukemia 1992; 6:31-6. hematopoiesis in sheep transplanted in utero during early 2 9 . Silvestri F, Banavali S, Baccarani M, Preisler HD. Progenitor gestation with fractionated adult human bone marrow cells. cell associated antigen CD34: biology and clinical applica- Blood 1992; 79:1404-12. tions. Haematologica 1992; 77:265-72. 4 5 . Berenson RJ, Bensinger WI, Kalamasz D, et al. Engraftment 3 0 . Peschle CH, Köller U. Cluster report: CD34. In: Knapp W, of dogs with Ia-positive marrow cells isolated by Avidin- Dörken B, Gilks WR, Rieber EP, Schmidt RE, Stein H, von Biotin immunoadsorption. Blood 1987; 69:1363-7. dem Borne AEGKr, eds. Leukocyte typing IV: white cell dif- 4 6 . Berenson RJ, Andrews RG, Bensinger WI, et al. Antigen ferentiation antigens. Oxford:Oxford University Press, 1989: C D 3 4 + marrow cells engraft lethally irradiated baboons. J 8 1 7 - 8 . Clin Invest 1988; 81:951-5. 3 1 . Civin CI, Trishmann TM, Fackler MJ, et al. Report on the 4 7 . Gore SD, Kastan MB, Civin CI. Normal human bone marrow CD34 cluster workshop. In: Knapp W, Dörken B, Gilks WR, precursors that express terminal deoxynucleotidyl transferase Rieber EP, Schmidt RE, Stein H, von dem Borne AEGKr, eds. include T-cell precursors and possible lymphoid stem cells. Leukocyte typing IV: white cell differentiation antigens. Blood 1991; 77:1681-90. Oxford:Oxford University Press, 1989: 818-25. 4 8 . Miller JS, Verfaillie C, McGlave P. The generation of human 3 2 . Lanza F, Moretti S, Papa S, Malavasi F, Castoldi G. Report on natural killer cells from CD34+/ D R – primitive progenitors in the fifth international workshop on human leukocyte differ- long term bone marrow culture. Blood 1992; 80:2182-7. entiation antigens, Boston, November 3-7, 1993. Haematolo- 4 9 . Loken MR, Civin CI, Bighee WL, Langlois RG, Jensen RH. gica 1994; 79:374-86. Coordinate glycosylation and cell surface expression of gly- 3 3 . Greaves MF, Colman SM, Bühring HJ, et al. Report on the cophorin A during normal human erythropoiesis. Blood CD34 cluster workshop. In: Schlossman SF , Boumsell L, 1987; 70:1959-61. Gilks W, Harlan JM, Kishimoto T, Morimoto C, Ritz J, Shaw 5 0 . Shaw VO, Civin CI, Loken MR. Flow cytometric analysis of S, Silverstein RL, Springer TA, Tedder TF, Tood RF, eds. human bone marrow. IV. Differential quantitative expression Leukocyte typing V: White cell differentiation antigens. of T-200 common leukocyte antigen during normal hemato- Oxford:Oxford University Press, 1995, in press. poiesis. J Immunol 1988; 140:1861-7. 3 4 . Sutherland DR, Marsh JCW, Davidson J, Baker MA, Keating 5 1 . Loken MR, Shaw VO, Dattilio K, Civin CI. Flow cytometric A, Mellors A. Differential sensitivity of CD34 epitopes to analysis of human bone marrow. I. Normal erythroid devel- cleavage by Pasteurella haemolytica glycoprotease: implica- opment. Blood 1987; 69:255-63. tions for purification of CD34-positive progenitor cells. Exp 5 2 . Sawada K, Krantz SB, Kans JS, et al. Purification of human Hematol 1992; 20:590-9. colony-forming units-erythroid and demonstration of specif- 3 5 . Lanza F, Castoldi GL. Large scale enrichment of CD34+ c e l l s ic binding of erythropoietin. J Clin Invest 1987; 80:357-66. by Percoll density gradients: a CML-based study design. In: 5 3 . Caux C, Dezutter-Dambuyant C, Schmitt D, Banchereau J. Wunder E, Serke S, Solovat H, Henon P, eds. Hematopoietic GM-CSF and TNF-a cooperate in the generation of dendritic stem cells. Dayton:AlphaMed Press, 1993:255-70. Langherans cells. Nature 1992; 360:258-61. 3 6 . Lanza F, Bi S, Castoldi GL, Goldman JM. Abnormal expres- 5 4 . Terstappen LWMM, Huang S, Safford M, Lansdorp P, Loken sion of N-CAM (CD56) adhesion molecule on myeloid and MR. Sequential generation of hematopoietic colonies derived progenitor cells from chronic myeloid leukemia. Leukemia from single nonlineage-committed CD34+/ C D 3 8 – p r o g e n i t o r 1993; 7:1570-5. cells. Blood 1991; 77:1218-27. 3 7 . Lanza F, Rigolin GM, Moretti S, Latorraca A, Castoldi GL. 5 5 . De Fabritiis P, Dowding C, Bungey J, et al. Phenotypic char- Prognostic value of immunophenotypic characteristic of acterization of normal and CML CD34-positive cells: only blasts cells in . Leuk Lymphoma 1994; the most primitive CML progenitors include Ph-neg cells. 13(Suppl 1):81-5. Leuk Lymphoma 1993; 11:51-61. 3 8 . Morecki S, Topalian SL, Myers WW, Okrongly D, Okarma 5 6 . Verfaillie C, Miller WJ, Boylan K, McGlave PB. Selection of CD34-positive cells 385

benign primitive hematopoietic progenitors in chronic myel- 7 7 . Leary AG, Zeng HQ, Clark SC, Ogawa M. Growth factor ogenous leukemia on the basis of HLA-DR antigen expres- requirements for survival in G0 and entry into the cell cycle sion. Blood 1992; 79:1003-10. of primitive human hematopoietic progenitors. Proc Natl 5 7 . Beschoner WE, Civin CI, Strauss LC. Localization of hemo- Acad Sci USA 1992; 89:4013-7. poietic progenitor cells in tissue with the anti-My-10 mono- 7 8 . Moore MAS. Clinical implications of positive and negative clonal antibody. Am J Pathol 1985; 119:1-6. hematopoietic stem cell regulators. Blood 1991; 78:1-19. 5 8 . Fina L, Molgaard HV, Robertson D, et al. Expression of the 7 9 . Hyrayama F, Shih JP, Awgulewitsch A, Warr GW, Clark SC, CD34 gene in vascular endothelial cells. Blood 1990; 75: Ogawa M. Clonal proliferation of murine lymphohematopoi- 2 4 1 7 - 2 6 . etic progenitors in culture. Proc Natl Acad Sci USA 1992; 89: 5 9 . Delia D, Lampugnani MG, Resnati M, et al. CD34 expression 5 9 0 7 - 1 1 . is regulated reciprocally with adhesion molecules in vascular 8 0 . Gordon MY, Riley GP, Watt SM, Greaves MF. Compart- endothelial cells in vitro. Blood 1993; 81:1001-8. imentalization of a haematopoietic growth factor (GM-CSF) 6 0 . Majdic O, Johannes Stockl, Pickl WF, et al. Signaling and by glycosaminoglycans in the bone marrow microenviron- induction of enhanced cytoadhesiveness via the hematopoiet- ment. Nature 1987; 326:403-5. ic progenitor cell surface molecule CD34. Blood 1994; 83: 8 1 . Montagna C, Massaro P, Morali F, Foa P, Maiolo AT, Eridani 1 2 2 6 - 3 4 . S. In vitro sensitivity of human erythroid progenitors to 6 1 . Borowitz MJ, Gockerman JP, Moore JO, et al. Clinico- hemopoietic growth factors: studies on primary and sec- pathologic and cytogenetic features of CD34 (MY10)-positive ondary polycythemia. Haematologica 1994; 79:311-8. acute nonlymphocytic leukemia. Am J Clin Pathol 1989; 8 2 . Jacobsen SEW, Ruscetti FW, Dubois CM, Wine J, Keller JR. 9 1 : 2 6 5 - 7 0 . Induction of colony-stimulating factor receptor expression 6 2 . Campos L, Guyotat D, Archimbaud E, et al. Surface marker on hematopoietic progenitor cells: Proposed mechanism for expression in adults with acute myelocytic leukemia: correla- growth factor synergism. Blood 1992; 80:678-87. tions with initial characteristics, morphology and response to 8 3 . Testa U, Pelosi E, Gabbianelli M, et al. Cascade of transacti- therapy. Br J Haematol 1989; 72:161-6. vation of growth factor receptors in early human hemato- 6 3 . Geller RB, Zahurak M, Hurwitz CA, et al. Prognostic impor- poiesis. Blood 1993; 81:1442-56. tance of immunophenotyping in adults with acute myelocytic 8 4 . Hirano T, Yasukawa K, Harada H, et al. Complementary leukaemia: the significance of the stem-cell DNA for a novel human interleukin (BSF-2) that induces B CD34 (My10). Br J Haematol 1990; 76:340-7. lymphocytes to produce immunoglobulin. Nature 1986; 324: 6 4 . Myint H, Lucie NP. The prognostic significance of the CD34 7 3 - 6 . antigen in acute myeloid leukaemia. Leuk Lymphoma 1992; 8 5 . Kitamura T, Sato N, Arai K, Miyajima A. Expression cloning 7 : 4 2 5 - 9 . of the human IL-3 receptor cDNA reveals a shared b subunit 6 5 . Lauria F, Raspadori D, Ventura MA, et al. CD7 expression for the human IL-3 and GM-CSF receptors. Cell 1991; 66: does not affect the prognosis in acute myeloid leukemia. 1 1 6 5 - 9 . Blood 1994; 83:3097-8. 8 6 . Yin T, Taga T, Lik-Shing Tsang M, Yasukawa K, Kishimoto 6 6 . Borowitz Mj, Shuster JJ, Civin CI, et al. Prognostic signifi- T, Yang YC. Involvement of IL-6 signal transducer gp130 in cance of CD34 expression in childhood B-precursor acute IL-11-mediated signal transduction. J Immunol 1993; 151: lymphocytic leukemia: a Pediatric Oncology Group Study. J 2 5 5 5 - 6 1 . Clin Oncol 1990; 8:1389-96. 8 7 . Massague J. The transforming growth factor-family. Ann Rev 6 7 . Pui CH, Hancock ML, Head DR, et al. Clinical significance of Cell Biol 1990; 6:597-641. CD34 expression in childhood acute lymphoblastic leukemia. 8 8 . Ploemacher RE, van Soest PL, Boudewijn A. Autocrine trans- Blood 1993; 82:889-98. forming growth factor 1 blocks colony formation and prog- 6 8 . Till JE, McCulloch EA. Hemopoietic stem cell differentiation. enitor cell generation by hemopoietic stem cells stimulated Biochim Biophys Acta 1980; 605:431-59. with steel factor. Stem Cells 1993; 11:336-47. 6 9 . Van Zant G, Chen JJ, Scott-Micus K. Developmental poten- 8 9 . Jacobsen SEW, Keller JR, Ruscetti FW, Kondaiah P, Roberts tial of hematopoietic stem cells determined using retrovirally AB, Falk LA. Bidirectional effects of transforming growth fac- marked allogenic marrow. Blood 1991; 77:756-63. tor b ( T G F -b) on colony-stimulating factor-induced human 7 0 . Till JE, McCulloch EA, Siminovitc L. A stochastic model of myelopoiesis in vitro: differential effects of distinct TGF- iso- stem cell proliferation, based on the growth of spleen colony- forms. Blood 1991; 78:2239-47. forming cells. Proc Natl Acad Sci USA 1964; 51:29-34. 9 0 . Keller JR, Jacobsen SEW, Dubois CM, Hestdal K, Ruscetti 7 1 . Ogawa M. Differentiation and proliferation of hematopoietic FW. Transforming growth factor-a: a bidirectional regulator stem cells. Blood 1993; 81:2844-53. of hematopoietic cell growth. Int J Cell Cloning 1992; 10:2-11. 7 2 . Bergamaschi G, Rosti V, Danova M, Lucotti C, Cazzola M. 9 1 . Lemoli RM, Fogli M, Fortuna A, Tura S. Interleukin-11 (IL- Apoptosis: biological and clinical aspects. Haematologica 11) and IL-9 counteract the inhibitory activity of transform- 1994; 79:86-93. ing growth factor b3 (TGF b3) on human primitive hemo- 7 3 . Barosi G. Control of erythropoietin production in man. Hae- poietic progenitor cells. Haematologica 1995; 80:5-12. matologica 1993; 78:77-9. 9 2 . Jacobsen SEW, Ruscetti FW, Dubois CM, Lee J, Boone TC, 7 4 . Cazzola M. The end of a long search: at last thrombopoietin. Keller JR. Transforming growth factor-b transmodulates the Haematologica 1994; 79:397-9. expression of colony stimulating factor receptors on murine 7 5 . Ikebuchi K, Clark SC, Ihle JN, Souza LM, Ogawa M. Granu- hematopoietic progenitor cell lines. Blood 1991; 77:1706-16. locyte colony-stimulating factor enhances interleukin-3- 9 3 . Laiho M, Di Caprio JA, Ludlow JW, Livingstone DM, Massa- dependent proliferation of multipotential hemopoietic prog- gue J. Growth inhibition by TGF-beta linked to suppression enitors. Proc Natl Acad Sci USA 1988; 85:3445-9. of retinoblastoma protein phosphorylation. Cell 1990; 62: 7 6 . Sonoda Y, Yang YC, Wong GG, Clark SC, Ogawa M. Analysis 1 7 5 - 8 5 . in serum-free culture of the targets of recombinant human 9 4 . Takehara K, Le Roy EC, Grotendorst GR. TGF-b i n h i b i t i o n hemopoietic factors: interleukin-3 and granulocyte/ma- of endothelial cell proliferation: alteration of EGF binding crophage colony-stimulating factor are specific for early and EGF-induced growth-regulatory (competence) gene developmental stages. Proc Natl Acad Sci USA 1988; 85: expression. Cell 1987; 49:415-22. 4 3 6 0 - 6 . 9 5 . Strife A, Lambek C, Perez A, et al. The effects of TGF-b3 on 386 C. Carlo Stella et al.

the growth of highly enriched hematopoietic progenitor cells 1 1 2 . Teshima T, Harada M, Takamatsu Y, et al. Cytotoxic drug derived from normal human bone marrow and peripheral and cytotoxic drug/G-CSF mobilization of peripheral blood blood. Cancer Res 1991; 15:4828-36. stem cells and their use for autografting. Bone Marrow 9 6 . Lemoli RM, Strife A, Clarkson BD, Haley JD, Gulati SC. Transpl 1992; 10:215-20. T G F -b3 protects normal human hematopoietic progenitor 1 1 3 . Siena S, Bregni M, Brando B, Ravagnani F, Bonadonna G, cells treated with 4-hydroperoxycyclophosphamide in vitro. Gianni AM. Circulation of CD34-positive hematopoietic Exp Hematol 1992; 20:1252-6. stem cells in the peripheral blood of high-dose cyclophos- 9 7 . Dubois CM, Ruscetti FW, Stankowa J, Keller JR. Trans- phamide treated patients: enhancement by intravenous forming growth factor-b regulates c-kit message stability and recombinant human GM-CSF. Blood 1989; 74:1905-14. cell-surface protein expression in hematopoietic progenitors. 1 1 4 . Ravagnani F, Siena S, Bregni M, Sciorelli G, Gianni AM, Blood 1994; 83:3138-45. Pellegris G. Large scale collection of circulating haematopoi- 9 8 . Caux C, Saeland S, Favre C, Duvert V, Mannoni P, Banche- etic progenitors in cancer patients treated with high-dose rau J. Tumor necrosis factor-a strongly potentiates inter- cyclophosphamide and recombinant human GM-CSF. Eur J leukin-3 and granulocyte-macrophage colony-stimulating Cancer 1990; 26:562-4. factor-induced proliferation of human CD34+ h e m a t o p o i e t i c 1 1 5 . Kawano Y, Takaue Y, Watanabe T, et al. Effects of progenitor progenitor cells. Blood 1990; 75:2292-8. cell dose and preleukapheresis use of human recombinant 9 9 . Schall TJ, Lewis M, Koller KJ, et al. Molecular cloning and granulocyte colony-stimulating factor on the recovery of expression of a receptor for human tumor necrosis factor. hematopoiesis after blood stem cell autografting in children. Cell 1990; 61:361-4. Exp Hematol 1993; 21:103-8. 1 0 0 . Rusten LS, Jacobsen FW, Lesslauer W, Loetscher H, Smeland 1 1 6 . Bensinger W, Singer J, Appelbaum F, et al. Autologous trans- EB, Jacobsen SEW. Bifunctional effects of tumor necrosis fac- plantation with peripheral blood mononuclear cells collected t o r - a ( T N F - a) on the growth of mature and primitive after administration of recombinant granulocyte stimulating human hematopoietic progenitor cells: involvement of p55 factor. Blood 1993; 81:3158-63. and p75 TNF receptors. Blood 1994; 83:3152-9. 1 1 7 . Haas R, Mohle R, Fruhauf S, et al. Patient characteristics 1 0 1 . Graham GJ, Wright EG, Hewick R, et al. Identification and associated with successful mobilizing and autografting of characterization of an inhibitor of hematopoietic stem cell peripheral blood progenitor cells in malignant lymphoma. proliferation. Nature 1990; 344:442-4. Blood 1994; 83:3787-94. 1 0 2 . Broxmeyer HE, Sherry B, Lu L, et al. Enhancing and sup- 1 1 8 . To LB, Dyson PG, Juttner CA. Cell-dose effect in circulating pressing effects of recombinant murine macrophage infla m- stem-cell autografting. Lancet 1986; ii:404-5. matory proteins on colony formation in vitro by bone mar- 1 1 9 . Gianni AM, Tarella C, Siena S, et al. Durable and complete row myeloid progenitor cells. Blood 1992; 76:1110-6. hematopoietic reconstitution after autografting of rhGM- 1 0 3 . Dunlop DJ, Wright EG, Lorimore S, et al. Demonstration of CSF exposed peripheral blood progenitor cells. Bone Marrow stem cell inhibition and myeloprotective effects of SCI/MIP Transplant 1991; 6:143-5. 1a in vivo. Blood 1992; 79:2221-5. 1 2 0 . Hohaus S, Goldschmidt H, Ehrhardt R, Haas R. Successful 1 0 4 . Moreb J, Zucali JR, Rueth S. The effects of tumor necrosis autografting following myeloablative conditioning therapy f a c t o r -a on early human hematopoietic progenitor cells with blood stem cells mobilized by chemotherapy plus rhG- treated with 4-hydroperoxycyclophosphamide. Blood 1990; CSF. Exp Hematol 1993; 21:508-14. 7 6 : 6 8 1 - 9 . 1 2 1 . Siena S, Bregni M, Di Nicola M, et al. Durability of hema- 1 0 5 . Andrews RG, Singer JW, Bernstein ID. Precursors of colony topoiesis following myeloablative cancer tharapy and auto- forming cells in human can be distinguished from colony grafting with peripheral blood hematopoietic progenitors. forming cells by expression of the CD33 and CD34 antigens Ann Oncol 1994; 5:935-41. and light scatter properties. J Exp Med 1989; 169:1721-31. 1 2 2 . Tarella C, Boccadoro M, Omedè P, et al. Role of chemothera- 1 0 6 . Gabbianelli M, Sargiacomo M, Pelosi E, Testa U, Isacchi G, py and GM-CSF on hemopoietic progenitor cell mobilization Peschle C. Pure human hematopoietic progenitors: permis- in multiple myeloma. Bone Marrow Transplant 1993; 11: sive action of basic fibroblast growth factor. Science 1990; 2 7 1 - 7 . 249:1561-4. 1 2 3 . Tarella C, Ferrero D, Siena S, et al. Conditions influ e n c i n g 1 0 7 . Peters W, Rosner G, Ross M, et al. Comparative effects of the expansion of the circulating hemopoietic progenitor cell granulocyte-macrophage colony-stimulating factor (GM- compartment. Haematologica 1990; 75:11-4. CSF) and granulocyte colony-stimulating factor (G-CSF) on 1 2 4 . Gianni AM, Bregni M, Siena S, et al. Granulocyte-macro- priming peripheral blood progenitor cells for use with autol- phage colony stimulating factor or granulocyte-colony stimu- ogous bone marrow transplantation after high-dose chemo- lating factor infusion makes high-dose etoposide a safe out- therapy. Blood 1993; 81:1709-19. patient regimen that is effective in lymphoma and myeloma 1 0 8 . Duhrsen U, Villeval JL, Boyd J, Kannourakis G, Morstyn G, patients. J Clin Oncol 1992; 10:1955-62. Metcalf D. Effects of recombinant human granulocyte 1 2 5 . Dreyfus F, Leblond V, Belanger C, et al. Peripheral blood colony-stimulating factor on hematopoietic progenitor cells stem cell collection and autografting in high risk lymphoma. in cancer patients. Blood 1988; 72:2074-81. Bone Marrow Transpl 1992; 10:409-13. 1 0 9 . Socinski MA, Elias A, Schnipper L, Cannistra SA, Antman 1 2 6 . Shimazaki C, Oku N, Ashihara E, et al. Collection of periph- KH, Griffin JD. Granulocyte-macrophage colony-stimulating eral blood stem cells mobilized by high-dose Ara-C plus VP- factor expands the circulating haemopoietic progenitor cell 16 or aclarubicin followed by recombinant human granulo- compartment in man. Lancet 1988; i:1194-8. cyte-colony stimulating factor. Bone Marrow Transpl 1992; 1 1 0 . Gianni AM, Siena S, Bregni M, et al. Granulocyte-macro- 10:341-6. phage colony-stimulating factor to harvest circulating 1 2 7 . Pettengell R, Testa NG, Swindell R, Crowther D, Dexter TM. haematopoietic stem cells for autotransplantation. Lancet Transplantation potential of hematopoietic cell released into 1989; ii:580-5. the circulation during routine chemotherapy for non- 1 1 1 . Kotasek D, Shepherd KM, Sage RE, et al. Factors affecting Hodgkin’s lymphoma. Blood 1993; 82:2239-48. blood stem cell collections following high-dose cyclophos- 1 2 8 . Tarella C, Ferrero D, Bregni M, et al. Peripheral blood expan- phamide mobilization in lymphoma, myeloma and solid sion of early progenitor cells after high-dose cyclophos- tumors. Bone Marrow Transplant 1992; 9:11-7. phamide and rhGM-CSF. Eur J Cancer 1991; 27:22-7. CD34-positive cells 387

1 2 9 . Bender JG, Lum L, Unverzagt KL, et al. Correlation of 7 7 : 4 0 0 - 9 . colony-forming cells, long-term culture initiating cells and 1 4 3 . Menichella G, Pierelli L, Foddai ML, et al. Autologous blood C D 3 4 + cells in apheresis products from patients mobilized for stem cell harvesting and transplantation in patients with peripharal blood progenitors with different regimens. Bone advanced ovarian cancer. Br J Haematol 1991; 79:444-50. Marrow Transpl 1994; 13:479-85. 1 4 4 . Pettengell R, Morgenstern GR, Woll PJ, et al. Peripheral 1 3 0 . Haas R, Ho AD, Bredthauer U, et al. Successful autologous blood progenitor cell transplantation in lymphoma and transplantation of blood stem cells mobilized with recombi- leukemia using a single apheresis. Blood 1993; 82:3770-7. nant human granulocyte-macrophage colony-stimulating 1 4 5 . Caracciolo D, Gavarotti P, Aglietta M, et al. High-dose factor. Exp Hematol 1990; 18:94-8. sequential (HDS) chemotherapy with blood and marrow cell 1 3 1 . Sheridan WP, Begley CG, Juttner C, et al. Effect of peripheral autograft as salvage treatment in very poor prognosis, blood progenitor cells mobilized by filgrastim (G-CSF) on relapsed non-Hodgkin lymphoma. Bone Marrow Transplant platelet recovery after high-dose chemotherapy. Lancet 1992; 1993; 12:621-5. i : 6 4 0 - 4 . 1 4 6 . Mayani H, Dragowska W, Lansdorp PM. Cytokine-induced 1 3 2 . Weaver CH, Buckner CD, Longin K, et al. Syngeneic trans- selective expansion and maturation of erythroid versus plantation with peripheral blood mononuclear cells collected myeloid progenitors from purified cord blood precursor after the administration of recombinant human granulocyte cells. Blood 1993; 81:3252-8. colony-stimulating factor. Blood 1993; 82:1981-4. 1 4 7 . Traycott CM, Abboud MR, Laver J, Clapp DW, Srour EF. 1 3 3 . Matsunaga T, Sakamaki S, Ohi S, Hirayama Y, Niitsu Y. Rapid exit from G0-G1 phases of cell cycle in response to Recombinant human granulocyte colony-stimulating factor stem cell factor confers on umbilical cord blood CD34+ c e l l s can mobilize sufficient amounts of peripheral blood stem an enhanced ex vivo expansion potential. Exp Hematol 1994; cells in healthy volunteers for allogeneic transplantation. 2 2 : 1 2 6 4 - 7 2 . Bone Marrow Transpl 1993; 11:103-8. 1 4 8 . Molineux G, Pojda Z, Hampson IN, Lord BI, Dexter TM. 1 3 4 . Schwinger W, Mache Ch, Urban Ch, Beaufort F, Toglhofer Transplantation potential of peripheral blood stem cells W. Single dose of filgrastim (rhG-CSF) increases the number induced by granulocyte colony-stimulating factor. Blood of hematopoietic progenitors in the peripheral blood of adult 1990; 76:2153-8. volunteers. Bone Marrow Transpl 1993; 11:489-92. 1 4 9 . Gianni AM, Bregni M, Siena S, et al. Rapid and complete 1 3 5 . Brugger W, Bross K, Frisch J, et al. Mobilization of peripheral hemopoietic reconstitution following combined transplanta- blood progenitor cells by sequential administration of tion of autologous blood and bone marrow cells. A changing Interleukin-3 and granulocyte-macrophage colony-stimulat- role for high dose chemo-radiotherapy? Hematol Oncol ing factor following polychemotherapy with etoposide, ifos- 1989; 7:139-48. famide and cisplatin. Blood 1992; 79:1193-200. 1 5 0 . Jones RJ, Wagner JE, Celano P, Zicha MS, Sharkis SJ. 1 3 6 . D’Hondt V, Guillaume T, Humblet Y, et al. Tolerance of Separation of pluripotent hematopoietic stem cells from sequential or simultaneous administration of IL-3 and G-CSF spleen colony forming cells. Nature 1990; 347:188-9. in improving peripheral blood stem cell harvesting following 1 5 1 . Uchida N, Aguila HL, Fleming WH, Jerabek L, Weissman IL. multi-agent chemotherapy: a pilot study. Bone Marrow Rapid and sustained hematopoietic recovery in lethally irra- Transpl 1994; 13:261-4. diated mice transplanted with purified Thy-1.1loLin-Sca-1+ 1 3 7 . de Revel T, Appelbaum FR, Storb R, et al. Effects of granulo- hematopoietic stem cells. Blood 1994; 83:3758-79. cyte colony-stimulating factor and stem cell factor, alone and 1 5 2 . Bradford G, Williams N, Barber L, Bertoncello I. Temporal in combination, on the mobilization of peripheral blood cells thrombocytopenia after engraftment with defined stem cells that engraft lethally irradiated dogs. Blood 1994; 83:3795-9. with long-term marrow reconstitution activity. Exp Hematol 1 3 8 . Liu KY, Akashi K, Harada M, Takamatsu Y, Niho Y. Kinetics 1993; 21:1615-20. of circulating haematopoietic progenitors during chemother- 1 5 3 . Robertson MJ, Soiffer RJ, Freedman AS, et al. Human bone apy-induced mobilization with or without granulocyte marrow depleted of CD33-positive cells mediates delayed but colony-stimulating factor. Br J Haematol 1993; 84:31-8. durable reconstitution of hematopoiesis: clinical trial of MY9 1 3 9 . Bishop MR, Anderson JR, Jackson JD, et al. High-dose thera- monoclonal antibody-purged autografts for the treatment of py and peripheral blood progenitor cell transplantation: acute myeloid leukemia. Blood 1992; 79:2229-36. effects of recombinant human granulocyte-macrophage 1 5 4 . Dreger P, Suttorp M, Haferlach T, et al. Allogeneic granulo- colony-stimulating factor on the autograft. Blood 1994; 83: cyte colony-stimulating factor-mobilized peripheral blood 6 1 0 - 6 . progenitor cells for treatment of engraftment failure after 1 4 0 . Jansen WE. Peripheral blood and bone marrow hematopoiet- bone marrow transplantation. Blood 1993; 81:401-9. ic stem cells: are they the same? Semin Oncol 1993; 20:19-27. 1 5 5 . Brenner MK, Rill DR, Moen RC, et al. Gene marking to trace 1 4 1 . Kessinger A, Armitage JO. The evolving role of autologous origin of relapse after autologous bone-marrow transplanta- peripheral stem cell transplantation following high-dose ther- tion. Lancet 1993; i: 85-6. apy for malignancies. Blood 1991; 77:211-3. 1 5 6 . Bregni M, Magni M, Siena S, Di Nicola M, Bonadonna G, 1 4 2 . Siena S, Bregni M, Brando B, et al. Flow cytometry for clini- Gianni AM. Human peripheral blood hematopoietic progen- cal estimation of circulating hematopoietic progenitors for itors are optimal targets of retroviral mediated gene transfer. autologous transplantation in cancer patients. Blood 1992; Blood 1992; 80:1418-22.