JOURNAL OF HEMATOTHERAPY 2:221-224 (1993)Mary Ann Liebert, Inc., Publishers

Ex Vivo Expansion and Gene Therapy Using Cord BloodCD34+ Cells

MALCOLM A.S. MOORE

ABSTRACT

CD34 * cells can be efficiently isolated from cord blood by sequential Percoll, Ficoll-Hypaque, andimmunomagnetic bead separation. This population, containing up to 25% progenitor cells and0.2-1% stem cells, defined by a long-term culture-initiating cell (LTC-initiating cell) assay, can beexpanded in vitro in the presence of a specific combination of cytokines resulting in a > 100,000-foldincrease in cells, a >2,000-fold increase in progenitors, and an 18- to 20-fold increase in LTC-initiating cells. These latter populations can be efficiently transfected with a retroviral vectorexpressing a mutated dehydrofolate-resistance gene that confers methotrexate reductase.

STIMULATION OF COLONY FORMATION of CD34+ Cellsseparated from cord blood using Percoll, Ficoll-

Hypaque, and immunomagnetic beads can best be accom-

plished using multiple cytokines. Clonogenic efficiencyof between 17 and 25% of the total CD34+ population canbe accomplished using combinations including fo'f-ligand(KL, stem cell factor), interleukin-3 (IL-3), interleukin-1(IL-1), interleukin-6 (IL-6), and erythropoietin (EPO)(Table 1).

When comparing the Ficoll-Hypaque-separated mono-

nuclear cell fractions of cord blood and adult peripheralblood leukapheresis preparations, a substantial number ofcolony-forming cells per 108 mononuclear cells are pro-duced in response to granulocyte-macrophage colony-stimulating factor (GM-CSF) (3 x 104-13 x 104). How-ever, when the CD34+ cells are isolated, the fraction thatresponds to GM-CSF is extremely low, only between 2and 6% of the original population of myeloid progenitors.This extremely low yield is due to the lack of an accessoryCD34-negative population and to the use of a singlecytokine. Much better progenitor recovery is accom-

plished when using multiple cytokine stimulation, espe-cially with the addition of IL-1, IL-3, and KL (52-100%recovery).

CD34+ cells from cord blood, from adult peripheralblood, and from normal adult bone marrow were used to

study their ability to expand in an in vitro system called a

delta assay (Table 2). In this 14-day expansion study,cells were cultured at limiting dilutions in the presence ofKL, IL-1, and IL-3. Even though the cord blood hadfewer progenitor cells relative to equivalent adult har-vests, the expansion potential was very substantial, ex-

ceeding that of adult marrow or adult peripheral bloodCD34+ cells.

Both total cells and total progenitor cells from anisolated CD34+ population of cord blood were exposed to

KL, IL-1, IL-6, and IL-3, with and without EPO (Table 3).In this ex vivo expansion system, amplification of bothpopulations peaked at about 21 days with a 105- to106-fold increase in total cells and a 3000- to 4000-foldincrease in progenitor cells. The addition of EPO did not

significantly influence the total expansion of myeloidprogenitors.

There were concerns that the ability to expand progen-itor cell populations might be at the expense of depletingstem cells. Attempts were made to quanti täte long-termculture (LTC)-initiating cells as an in vitro correlate ofearly stem cells. Measurements were based on cocultur-

SKI Cell Biology and Genetics Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021

221

Table 1. Colonies per 105 CD34+ Cells Stimulated by Cytokine Combinations3

IL-10±0

IL-31,903 ± 569

IL-60±0

KL149 ± 97

EPO0±0

IL-1 + KL227 ± 167

IL-3 + KL4,632 ± 1120

IL-6 + KL1,634 ± 1,546IL-3 + IL-6

2,900 ± 500

EPO + KL4,303 ± 1,111

IL-3 + IL-12,133 ±417

IL-1 + IL-3 + KL7,347 ± 1616

IL-6 + IL-3 + KL6,556 ± 1883

EPO + KL + IL-310,836 ± 3086

IL-1 + IL-3 + KL + EPO17,475 ± 4,268

IL-6 + IL-3 + KL + EPO15,350 ± 3783

a All assays at 103 cells/ml in agarose, scored on day 14. For all EPO-containing assays, colonycount is CFU-GM + BFU-E + CFU-GEMM. Data are from 3 to 10 separate experiments.

ing populations of cord blood-derived CD34+ cells thatwere expanded ex vivo for various periods of time. Thecells were then inoculated onto irradiated marrow stroma

(SV40-transformed stromal lines) and the number ofprogenitors generated after 5 to 8 weeks was measured.Between 0.2 and 1% of the CD34+ cells generatedprogenitors. This was considered the long-term culture-initiating cell frequency in the original cord blood popu-lation.

Various support matrices and stromal cell types that can

support the proliferation of LTC-initiating cells were

compared using cord blood-derived CD34+ populations.There is evidence to believe that marmoset bone marrow

fibroblasts are particularly effective at sustaining long-term hematopoietic stem cell renewal. These and SV40-transformed human fetal spleen and bone marrow stromalcell lines were compared with normal marrow stroma forthe support of cord blood LTC-initiating cells. The fre-quency of LTC-initiating cells was between 0.3 and 1 % ofthe CD34+ population when assayed on these stromalsources.

A major question is, can LTC-initiating cells be ex-

panded? LTC-initiating cells from cord blood can beexpanded at least 18- to 20-fold in suspension culture

Table 2. Comparison of "Double Delta" 14-DayExpansion of Marrow, Cord Blood, and

Cytoxan + G-CSF Elicited Peripheral Blood CD34+Cells with IL-1 + IL-3 + KL

Fold expansionCells Progenitors

MarrowPeripheral bloodCord blood

256

59 ±20223 ± 69

1103 ± 200

152 ± 40109 ± 19334 ± 22

containing KL, IL-1, and IL-3 over 2-3 weeks (Table 4).This indicates that the primitive cell population, which isprobably representative of the stem cell pool, is not lost inthe isolation and expansion process.

Substantial stochastic results were observed whenLTC-initiating cells were analyzed by isolating such cellsand then serially expanding them in culture over 5 to 7weeks. Some single LTC-initiating cells produced be-tween 109 and 1010 cells while others had considerablyreduced proliferative potential. Based on these findings itis important to recognize the fundamental randomness ofthe process of self-renewal and differentiation whenworking with limiting dilutions of stem cell populations invitro.

CORD BLOOD CD34+ CELLS AS TARGETSFOR THE INTRODUCTION OF GENETIC

INFORMATION USINGRETROVIRAL VECTORS

The retroviral vector we are using contains the gene fora mutated form of dihydrofolate reducíase that confersmethotrexate resistance to the hematopoietic population.This vector was introduced into the cord blood-derivedCD34+ population of cells. The transfected CD34+ cordcells were cocultured with a producing cell line, either indirect contact with the packaging cell line or separated bya cell-impermeable membrane that allows virion passagebut does not allow cells to migrate across the membrane.The cells were expanded in the presence of cytokinecombinations and selected with methotrexate. The cellsthat have been selected are presumably those to which a

relative resistance to methotrexate has been conferred.These cells can be expanded ex vivo. The conferredresistance is not absolute but the mutated gene shifts the in

222

GENE THERAPY USING CORD BLOOD CD34+ CELLS

Table 3. Cumulative Production of CFU-GM3 in Cultures Initiated with 4 x 104CD34+ Cells from Cord Blood and Passaged Weekly with 4 x io4 Cells

and Cytokines

Stimulus Peak CFU-GM number Fold increase DaysIL-6 + KL + IL-3IL-6 + KL + IL-3IL-6 + KL + IL-3 + EPObIL-1 + KL + IL-3IL-1 + KL + IL-3 + EPOIL-1 + KL + IL-6IL-1 + KL + IL-6 + EPO

1.39 x 1075.3 x 1062.7 x 1066.3 x 1064.1 x 1060.4 x 1060.5 x 106

3.73 x 1033.43 x 103

1.7 x 1032.8 x 1032.0 x 1030.3 x 1030.3 x 103

34 ± 0.32128

25 ± 125 ±4

2814

"Stimulus for CFU-GM, IL-1, or IL-6 + KL + IL-3 + EPO.bCocultured for the first week with mouse fibroblast line AM-12.

vitro progenitor cell dose response to methotrexate byapproximately one log of sensitivity.

CONCLUSIONS

These results show that cord blood-derived CD34+cells, which include stem cell populations, can be ex-

panded in vitro and are very effective targets for retroviraltransfection, opening the way for a number of strategies,including drug resistance and HIV-resistant gene trans-

fection.

QUESTIONS AND ANSWERS

Dr. Broxmeyer (Indianapolis): Would you repeat the 7,14, and 21 day results from the expansion of cord bloodprogenitor cells?Dr. Moore: In 7 days it is between 50- and 100-fold, by14 days almost 1,000-fold, and the maximum expansionat 21 to 28 days is about 10,000-fold.Dr. Broxmeyer: The numbers make a lot of sense

because we have done expansion with totally unseparated

cells and get at least a 10-fold increase after 7 days. It isnice to see you are not losing LTC-initiating cells.Dr. Moore: I think that is the take-home message—weare not losing the primitive cells.Dr. Harris (Tucson): What is your retroviral transfectionefficiency and are the LTC-initiating cells transfected?Dr. Moore: We do PCR analysis of individual coloniesand the transfection efficiency is between 60 and 90%introduction of the retroviral vector. That does not mean itis expressing the protein, but data from the methotrexateresistance, based on a lot of calculations and done-response analysis, would suggest that on the order of 60 to100% of the progenitors and LTC-initiating cells are

resistant.Dr. Harris: When you do the transfection after the initialexpansion, do you have that high a percentage of cells thatare in the cell cycle?Dr. Moore: Yes. I am using stem cell factor, IL-3 andIL-6. It is not just the proliferation, there is also some

significant up-regulation of receptors for the virus thatenhances our ability to transduce the gene.Dr. Gebig (Indianapolis): Do you think that the expan-sion of the LTC-initiating cells will correlate with the

Deltano.

Table 4. Expansion of Cord Blood CD34+ Cells in Weekly Delta CulturesStimulated with IL-1 + IL-3 + KL + EPOa

Cells x 106Fold

CFUIBFU x IO5Fold

increase LTC-ICFold

increase

0.043.5

110.61,5072,7784,036

188

2,80038,00069,000

101,000

0.032.2

19.073.210.03.2

172

6142,364

323104

3005,0005,530ND

11718

ND

aCFU/BFU stimulated by IL-1 + IL-3 + KL + EPO. ND, not detected. LTC-IC, long-term culture-initiating cells assayed at5 weeks on irradiated SV40-transformed human marrow fibroblasts.

223

MOORE

corresponding expansion of cells that will reconstitutehematopoiesis, if you had an animal model?Dr. Moore: We do have an animal model. We (Muenchet al, 1993) did the ex vivo expansion based on a

combination of SCF, IL-1, and IL-6, expanding the cellsfor 7 days and then using limiting dilutions of as few as

1000 cells. We were able to show rapid recovery ofplatelets and neutrophils in the short term. After a year,with both primary and secondary transplants and using Ychromosome markers, we convinced ourselves that wewere not depleting the stem cell long-term repopulatingunits by ex vivo expansion. I could not convince myselfthat I was expanding them. I would have convincedmyself that I had expanded them had I carried thetitrations to 100 or to 10 cells, but I carried the titrationsonly to 1000 cells.

REFERENCE

Muench, M.O., Firpo, M.T., & Moore, M.A.S. (1993): Bonemarrow transplantation with interleukin-1 plus kit-ligand ex

vivo expanded bone marrow accelerates hematopoietic recon-

stitution in mice without the loss of stem cell function. Blood81:3463.

Address reprint requests to:Malcolm A.S. Moore, D.Phil.

SKI Cell Biology and Genetics ProgramMemorial Sloan-Kettering Cancer Center

1275 York AvenueNew York, NY 10021

224