M. Hakki et al. / Biol Blood Marrow Transplant 20 (2014) 128e138132

presence of dysfunctional antigen-specific CD8þ T cells. Blood. 2002;100:3690-3697.

22. Omar H, Ahmed R, Rane L, et al. Decreased IL-7 signaling in T cells frompatients with PTLD after allogeneic HSCT. J Immunother. 2011;34:390-396.

23. Macchia I, Gauduin MC, Kaur A, Johnson RP. Expression of CD8alphaidentifies a distinct subset of effector memory CD4þ T lymphocytes.Immunology. 2006;119:232-242.

24. Oevermann L, Lang P, Feuchtinger T, et al. Immune reconstitution andstrategies for rebuilding the immune system after haploidentical stemcell transplantation. Ann N Y Acad Sci. 2012;1266:161-170.

Financial disclosure: See Acknowledgments on page 135.* Correspondence and reprint requests: Jay A. Nelson, Vaccine and Gene

Therapy Institute, Oregon Health and Sciences University, 505 NW 185thAve, Beaverton, OR 97006.

E-mail address: nelsonj@ohsu.edu (J.A. Nelson).1083-8791/$ e see front matter � 2014 American Society for Blood andMarrow Transplantation.http://dx.doi.org/10.1016/j.bbmt.2013.10.019

25. Scheper W, van Dorp S, Kersting S, et al. gdT cells elicited by CMVreactivation after allo-SCT cross-recognize CMV and leukemia. Leuke-mia. 2013;27:1328-1338.

26. Sinha ML, Fry TJ, Fowler DH, et al. Interleukin 7 worsens graft-versus-host disease. Blood. 2002;100:2642-2649.

27. Dean RM, Fry T, Mackall C, et al. Association of serum interleukin-7levels with the development of acute graft-versus-host disease. J ClinOncol. 2008;26:5735-5741.

28. Napolitano LA, Grant RM, Deeks SG, et al. Increased production of IL-7accompanies HIV-1-mediated T-cell depletion: Implications for T-cellhomeostasis. Nature Med. 2001;7:73-79.

HCMV Infection of Humanized Mice after Transplantationof G-CSFeMobilized Peripheral Blood Stem Cells fromHCMV-Seropositive Donors

Morgan Hakki 1, Devorah C. Goldman 2, Daniel N. Streblow3,Kimberly L. Hamlin 2, Craig N. Krekylwich 3, William H. Fleming 2,Jay A. Nelson 3,*

1Division of Infectious Diseases, Oregon Health and Science University, Portland, Oregon2Oregon Stem Cell Center, Department of Pediatrics, Oregon Health and Science University, Portland, Oregon3Vaccine and Gene Therapy Institute, Oregon Health and Science University, Beaverton, Oregon

Article history:

Received 28 August 2013Accepted 7 October 2013

Key Words:CytomegalovirusStem cell transplantationTransmissionLatencyNSG mouse model

a b s t r a c tHuman cytomegalovirus (HCMV) infection, including primary infection resulting from transmission from aseropositive donor to a seronegative recipient (Dþ/R�), remains a significant problem in the setting of pe-ripheral blood stem cell transplantation (PBSCT). The lack of a suitable animal model for studying HCMVtransmission after PBSCT is a major barrier to understanding this process and, consequently, developing novelinterventions to prevent HCMV infection. Our previous work demonstrated that human CD34þ progenitor celleengrafted NOD-scid IL2Rgc

null (NSG) mice support latent HCMV infection after direct inoculation and reac-tivation after treatment with granulocyte colony-stimulating factor. To more accurately recapitulate HCMVinfection in the Dþ/R� PBSCT setting, granulocyte colony-stimulating factoremobilized peripheral blood stemcells from seropositive donors were used to engraft NSG mice. All recipient mice demonstrated evidence ofHCMV infection in liver, spleen, and bone marrow. These findings validate the NSG mouse model for studyingHCMV transmission during PBSCT.

� 2014 American Society for Blood and Marrow Transplantation.

INTRODUCTIONDespite advances in diagnostics and therapeutics, human

cytomegalovirus (HCMV) remains a significant cause ofmorbidity andmortality after peripheral blood stem cell (PBSC)transplantation (PBSCT), and novel approaches to preventingHCMV infection are needed [1]. HCMV-seronegative recipients(R�) of allografts from HCMV-seropositive donors (Dþ),although at lower risk for developing HCMV infection anddisease than seropositive recipients (Rþ), will still developpost-transplantation primary infection in up to 20% of cases[2-7]. The donor graft is the most important source of virusearly in the post-transplantation period, and a retrospectiveanalysis of Dþ/R� transplantations identified several factorsassociatedwith successful virus transmission [5]. However, thestrict species specificity of cytomegaloviruses and the conse-quent lack of a suitable animal model system have hindered

the experimental validation of these findings as well as thedevelopment of preventative strategies in this population.

“Humanized” mice, which have undergone trans-plantation with human cells and/or tissues, have recentlybeen developed as tools to aid the in vivo study of pathogenswith strict human tropism [8]. We reported the first hu-manized mouse model of HCMV infection in which humanCD34þ hematopoietic progenitor celleengrafted NOD-scidIL2Rgc

null (NSG) mice directly infected with HCMV supportedlatent viral infection, reactivation in human macrophages,and dissemination after granulocyte colony-stimulatingfactor (G-CSF)-induced mobilization of bone marrow he-matopoietic cells [9]. The present study was carried out todetermine whether NSG mice would also demonstrateevidence of HCMV infection after transplantation of G-CSFemobilized PBSCs from HCMV-seropositive donors,thereby recapitulating Dþ/R� PBSCT and validating the NSGmouse model as a tool for studying HCMV transmission andinfection in this setting.

METHODSMice

NSG mice were maintained in an specific pathogen-free facility accord-ing to procedures approved by Oregon Health and Science University’sInstitutional Animal Care and Use Committee. Before transplantation, mice

Figure 1. Engraftment of NSG mice after transplantation of G-CSFemobilized PBSCs. (A) Representative donor lineage analysis in xenografts. Flow cytometry ofsplenocytes from a mouse undergoing transplantation with PBSCs from donor 2 is shown. (B-D) Peripheral blood (B), bone marrow (C), and splenocytes (D) frommiceundergoing transplantation with PBSCs from donors 2, 3, and 4 were analyzed at the time of tissue harvest for engraftment of human cell populations. Donor 1human CD45þ engraftment was <0.5% in all mice.

M. Hakki et al. / Biol Blood Marrow Transplant 20 (2014) 128e138 133

Table 1Evidence of HCMV Infection in NSG Mice after PBSCT from HCMV-Seropositive Donors

Donor/Recipient Viral Load*

Bone Marrow Liver Spleen

Donor 11 13,130 2060 14552 110,991 9350 83223 24,243 696 23554 11,162 16,809 612Average 39,881 7229 3186

Donor 21 18,443 419 2142 951 1013 25743 25,252 2100 90294 1170 2627 544Average 11,454 1540 3090

Donor 31 513 1240 1672 3039 352 923 2741 539 3494 863 2351 305Average 1789 1121 228

Donor 4y

1 141 121 02 106 185 03 134 46 100Average 127 118 33

* HCMV genomic copies per mg of DNA.y HCMV-seronegative donor.

M. Hakki et al. / Biol Blood Marrow Transplant 20 (2014) 128e138134

were sublethally irradiated (100-250 cGy) with a JL Shepherd Mark I 137Csirradiator (JL Shepherd Associates, San Fernando, CA) or a Rad Source RS2000 X-ray irradiator (Rad Source Technologies, Suwanee, GA). After trans-plantation, adult micewere providedwith antibiotic water (1.1 g/L neomycinsulfate and 167 mg/L polymyxin B) up to the time of tissue harvest.

Human Donor Cell TransplantationG-CSFemobilized PBSCs were obtained from 4 donors after informed

consent in accordance with Oregon Health and Science University Institu-tional Review Board policies. PBSCs were transplanted i.v. into the retro-orbital plexus of 8- to 12-wk-old NSG mice. Mice were sacrificed at 6-8wk after transplantation or sooner if they appeared sick. Peripheral blood,bone marrow, liver, and spleen were collected at the time of euthanasia.Peripheral blood, bone marrow, and a portion of the spleen were processedimmediately after euthanasia to obtain leukocytes for flow cytometry. Por-tions of each liver and spleen were placed in RNAlater (Life Technologies,Grand Island, NY) for preservation until processing for DNA. Human cellengraftment was assessed by flow cytometry analysis of adult mouse pe-ripheral blood as described previously [9].

Quantification of HCMV Genomic DNATotal DNA was extracted from tissue or isolated cells with DNAzol (Life

Technologies) following the manufacturer’s suggested protocol, with theaddition of a second DNA precipitation to remove any residual contamina-tion. In brief, 0.025 g of tissue was placed in an Eppendorf tube containing1 mL of DNAzol reagent with 1-mm glass beads. The samples were ho-mogenized for 3 minutes in a BeadBeater (Precellys 24, Bertin Technologies,Rockville, MD) tissue homogenizer. Total DNA was precipitated by theaddition of 100% ethanol (EtOH), followed by centrifugation at 16,000� g for10 min. The resulting DNA pellet was washed twice with 75% EtOH, resus-pended in distilled water, and precipitated a second time with the additionof sodium acetate and EtOH.

DNA was analyzed by quantitative real-time polymerase chain reaction(RT-PCR) using a primer and probe set directed against the HCMV US28 genesequence: probe, TGATCC CGC TCAGTGT; forward primer, GAACTC ATG CTCGGT GCT TTC; reverse primer, CTT TGT GGC GCG ACT GAG A. This primer/probe set was identified using Primer Express software (Applied Biosystems,Wilmington, DE). RT-PCR reactions were performed using Fast AdvancedMaster Mix (Applied Biosystems) under the following reaction conditions:thermal activation of AmpliTaq Gold (20 s at 95�C), followed by a total of 40cycles of PCR (1 s at 95�C and 20 s at 60�C) using the ABI Prism Step One Plussequence detection system (Applied Biosystems). PCR results were analyzedusing ABI Prism Step One sequence detection software. HCMV bacterialartificial chromosomal DNA served as a quantification standard. The sensi-tivity of detection of this assay was w50 HCMV genomic copies.

RESULTS AND DISCUSSIONThe objective of this study was to determine whether

transplantation of G-CSFemobilized PBSCs from HCMV-seropositive donors would result in successful infection ofrecipient NSG mice, similar to what occurs during trans-plantation of allografts from HCMV-seropositive donors to-seronegative recipients [5]. Archived G-CSFemobilizedPBSCs from 3 different HCMV-seropositive donors (donors1-3) were transplanted into a total of 12 sublethally irra-diated NSG mice (4 mice per donor). As controls, 3 NSGmice underwent transplantation with G-CSFemobilizedPBSCs from an HCMV-seronegative donor (donor 4). Allmice received approximately 2 � 106 unfractionated cellsi.v. and were evaluated for overall health daily aftertransplantation.

The cohort of mice that received PBSCs from donor 1weresacrificed at 1 wk after transplantation as they began toappear sickly, with weight loss and loss of normal fur,consistent with xenogenic graft-versus host disease (GVHD).Mice that received PBSCs from donors 2, 3, and 4 appearedwell until the planned harvest at 6, 8, and 6 wk after trans-plantation, respectively. Peripheral blood, spleen, and bonemarrow from eachmouse (except donor 1) were analyzed forengraftment by flow cytometry and for HCMV DNA byquantitative RT-PCR.

At the time of harvest at 6-8 wk after transplantation,mice that received PBSCs from donors 2-4 displayed signifi-cant levels of human CD45þ cells in the peripheral blood(mean, 5.5% � 5.5%), bone marrow (mean, 12.6% � 21.2%),and spleen (mean, 17.3% � 29.3%) (Figure 1). Although themajority of engrafted human CD45þ cells were either CD3þ Tcells or CD19þ B cells, significant numbers of CD14-expressing myelomonocytic cells were detected as well(mean, 8% � 14% for peripheral blood, 6.3% � 7.8% for bonemarrow, and 5.1% � 10.2% for spleen).

QuantitativeRT-PCR forHCMVDNAperformedonsamplesobtained from mice that underwent PBSCT from HCMV-seronegative donor 4 yielded low-level positive results, witha mean range of 33-127 copies/cell in the bone marrow, liver,and spleen (Table 1). These results likely reflect some degreeof nonspecific amplification of either human or, less likely,mouse DNA. Using the maximum quantity of HCMV detectedin any sample in the cohort ofmice undergoingPBSCT fromanHCMV-seronegative donor 4 (185 copies/mg of DNA) as abaseline, 11 of 12 NSGmice that underwent PBSCT from the 3seropositive donors had virus detected at >10-fold abovebaseline levels in bone marrow, liver, and spleen. Given thatHCMV will not replicate in these native mouse tissues, thisfinding most likely reflects the detection of disseminatedHCMVpresent in cells derived fromtransplantedhumanPBSCproduct. RT-PCR assays without template performed witheach PCR runwere consistently negative (0 copies), indicatingthat the results obtained could not not be attributed to cross-contaminationwithHCMVDNA.QuantitativeRT-PCRanalysisof HCMV DNA present in total stem cell product from eachdonorbefore transplantationwasunsuccessful, likelybecauseof the extremely low levels of virus present [10].

Based on our results, we conclude that G-CSFemobilizedPBSCs obtained from HCMV- seropositive donors represent asource of infectious virus that is capable of transmission anddissemination in the NSG mice model. This work comple-ments and expands on previous studies using NSGmice as anin vivo model of HCMV latency and reactivation [9]. Thismodel can be applied to determine the cellular reservoir oflatent virus in PBSCs that is necessary and sufficient for

T. Ruutu et al. / Biol Blood Marrow Transplant 20 (2014) 128e138 135

transmitting HCMV from donor to recipient, allowing the useof standard cell separation techniques to enrich or depletethe stem cell product of specific cell types before trans-plantation into NSG hosts. Of particular interest are myeloidlineage cells, such as CD34þ progenitor cells and CD14þ

monocytes, which have been previously described asharboring latent HCMV [11]. Defining the cell populationsthat harbor latent HCMV may lead to novel strategies toprevent transmission during PBSCT.

Why only 15%-20% of seropositive donors successfullytransmit HCMV to seronegative recipients is not clear. Theprevious lackof an experimentalmodel limited the analysis oftransmission to retrospective clinical studies. The NSGmousemodel now permits the experimental evaluation of factorsinherent to the allograft that have been found to correlatewith transmission in such studies [5]. In addition, this modelprovides a tool for testing novel hypotheses, such as whetherthe risk of transmission is related to the quantitative viral loadin the allograft. Thus, theNSGmousemodel provides a uniqueopportunity to gain further insight into the fundamentalmechanisms of HCMV transmission and latency after PBSCT.

ACKNOWLEDGMENTSFinancial disclosure: This work was supported by National

Institutes of Health research grants AI21640 (to J.A.N.) andHL069133 (to W.H.F.). M.H. was supported by a faculty devel-opment award from the Sunlin and Priscilla Chou Foundation.

Conflict of interest statement: There are no conflicts of in-terest to report.

Authorship statement: M.H., D.C.G., D.N.S., W.H.F., andJ.A.N. designed research; MH, D.C.G., D.N.S., K.L.H., and C.N.K.performed experiments; M.H., D.C.G., D.N.S., W.H.F., and

J.A.N. analyzed results; M.H. and D.C.G. created the tables andfigures; and M.H. wrote the manuscript.

REFERENCES1. Boeckh M, Geballe AP. Cytomegalovirus: pathogen, paradigm, and

puzzle. J Clin Invest. 2011;121:1673-1680.2. Boeckh M, Nichols WG. The impact of cytomegalovirus serostatus of

donor and recipient before hematopoietic stem cell transplantation inthe era of antiviral prophylaxis and preemptive therapy. Blood. 2004;103:2003-2008.

3. George B, Pati N, Gilroy N, et al. Pre-transplant cytomegalovirus (CMV)serostatus remains the most important determinant of CMV reac-tivation after allogeneic hematopoietic stem cell transplantation in theera of surveillance and preemptive therapy. Transpl Infect Dis. 2010;12:322-329.

4. Nichols WG, Corey L, Gooley T, et al. High risk of death due to bacterialand fungal infection among cytomegalovirus (CMV)-seronegative re-cipients of stem cell transplants from seropositive donors: evidence forindirect effects of primary CMV infection. J Infect Dis. 2002;185:273-282.

5. Pergam SA, Xie H, Sandhu R, et al. Efficiency and risk factors for CMVtransmission in seronegative hematopoietic stem cell recipients. BiolBlood Marrow Transplant. 2012;18:1391-1400.

6. Walker CM, van Burik JA, De For TE, Weisdorf DJ. Cytomegalovirusinfection after allogeneic transplantation: comparison of cord bloodwith peripheral blood and marrow graft sources. Biol Blood MarrowTransplant. 2007;13:1106-1115.

7. Matthes-Martin S, Lion T, Aberle SW, et al. Pre-emptive treatment ofCMV DNAemia in pediatric stem cell transplantation: the impact ofrecipient and donor CMV serostatus on the incidence of CMV diseaseand CMV-related mortality. Bone Marrow Transplant. 2003;31:803-808.

8. Legrand N, Ploss A, Balling R, et al. Humanized mice for modelinghuman infectious disease: challenges, progress, and outlook. Cell HostMicrobe. 2009;6:5-9.

9. Smith MS, Goldman DC, Bailey AS, et al. Granulocyte-colony stimu-lating factor reactivates human cytomegalovirus in a latently infectedhumanized mouse model. Cell Host Microbe. 2010;8:284-291.

10. Slobedman B, Mocarski ES. Quantitative analysis of latent humancytomegalovirus. J Virol. 1999;73:4806-4812.

11. Sinclair J. Human cytomegalovirus: latency and reactivation in themyeloid lineage. J Clin Virol. 2008;41:180-185.

Improved Survival with Ursodeoxycholic Acid Prophylaxisin Allogeneic Stem Cell Transplantation: Long-TermFollow-Up of a Randomized Study

Tapani Ruutu 1,*, Eeva Juvonen 1,2, Mats Remberger 3, Kari Remes 4,Liisa Volin 1, Jonas Mattsson 3, Anne Nihtinen 1, Hans Hägglund 3,Olle Ringdén 3, on behalf of the Nordic Group for Blood and MarrowTransplantation1Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland2 Finnish Red Cross Blood Transfusion Service, Helsinki, Finland3Centre for Allogeneic Stem Cell Transplantation, Departments of Clinical Immunology and Medicine, Karolinska Hospital, HuddingeUniversity Hospital, Huddinge, Sweden4Department of Medicine, Turku University Hospital, Turku, Finland

Article history:

Received 23 June 2013Accepted 15 October 2013

Key Words:Ursodeoxycholic acidUrsodiolAllogeneic stem celltransplantationTransplant-related mortalitySurvival

a b s t r a c tWe report the long-term results of a prospective randomized study on the use of ursodeoxycholic acid(UDCA) for prevention of hepatic complications after allogeneic stem cell transplantation. Two hundredforty-two patients, 232 with malignant disease, were randomized to receive (n ¼ 123) or not to receive(n ¼ 119) UDCA from the beginning of the conditioning until 90 days post-transplantation. The results werereported after 1-year follow-up. UDCA administration reduced significantly the proportion of patientsdeveloping high serum bilirubin levels as well as the incidence of severe acute graft-versus-host disease(GVHD), liver GVHD, and intestinal GVHD. In the UDCA prophylaxis group, nonrelapse mortality (NRM) waslower and overall survival better than in the control group. After a 10-year follow-up, the difference in thesurvival and NRM in favor of the UDCA-treated group, seen at 1 year, was maintained (survival 48% versus38%, P ¼ .037; NRM 28% versus 41%, P ¼ .01). A landmark analysis in patients surviving at 1 year post-transplantation showed no significant differences between the study groups in the long-term follow-up