specificities was not occurring. The kinet-

ics of T cells expressing this activated

phenotype also closely match those of

the antigen-specific T cells measured by

ICCS as well as by MHC Class I tetramer

staining. However, YFV-specific and VV-

specific memory T cells soon lose the

CD38+HLA-DR+ and the Ki-67+Bcl-2–

phenotype, and therefore these markers

are only useful for approximating the total

T cell response elicited during the acute

phase of infection. Nonetheless, this pro-

vides food for thought in terms of the real-

ization that the antiviral T cell response

might be larger than what we can measure

through current functional assays such as

IFNg-based ICCS.

There are several questions that remain

unanswered with regard to the study by

Miller et al. (2008). For instance, if intracel-

lular IFNg production ‘‘tags’’ only one-

third of the total virus-specific T cell

response (as measured by an activated

CD38+HLA-DR+ or Ki-67+Bcl-2– T cell

phenotype), then what cytokines, if any,

do the other two-thirds of the CD8+ T cells

make following stimulation with antigen?

Does this mean that the majority of the T

cell response represents non-IFNg pro-

ducers in humans, or is it possible that

our approaches to in vitro stimulation

require further optimization? Others have

commented that because of the multi-

functional nature of heterologous T cell

populations, measurement of any one

cytokine alone could lead to a lower,

more-conservative estimate of the total

antigen-specific T cell response (Seder

et al., 2008). With this in mind, it will be in-

teresting to learn the full spectrum of hu-

man cytokine production profiles that

are elicited following various acute viral

infections in both the effector and the

memory phases of the immune response.

On the basis of the results of the study

by Miller et al. (2008), it appears clear

that virus-specific T cell responses follow-

ing acute infection in humans are larger

than expected and vary substantially de-

pending on the efficiency of the technique

used to measure them. Importantly, both

YFV and VV represent viruses that infect

the host through a peripheral route. It will

be interesting to learn whether respiratory

viral infections such as influenza virus or

respiratory syncytial virus are also capa-

ble of eliciting large frequencies of virus-

specific CD8+ T cells in the bloodstream.

Further in-depth analysis of virus-specific

T cell kinetics will be important, given that

it appears that CD8+ T cell responses de-

cay very rapidly in the short term (i.e., the

first weeks after infection) compared to

the long-term (i.e., months or years after

vaccination [Co et al., 2002; Crotty et al.,

2003; Hammarlund et al., 2003]). Analysis

of T cell responses following booster

vaccination or re-infection will be another

interesting avenue of investigation. To-

gether, these studies will lead to a better

understanding of antiviral T cell respon-

ses and maintenance of immunological

memory in humans.

REFERENCES

Co, M.D., Terajima, M., Cruz, J., Ennis, F.A., andRothman, A.L. (2002). Virology 293, 151–163.

Crotty, S., Felgner, P., Davies, H., Glidewell, J.,Villarreal, L., and Ahmed, R. (2003). J. Immunol.171, 4969–4973.

Dasgupta, A., Hammarlund, E., Slifka, M.K., andFruh, K. (2007). J. Immunol. 178, 1654–1661.

Hammarlund, E., Lewis, M.W., Hansen, S.G.,Strelow, L.I., Nelson, J.A., Sexton, G.J., Hanifin,J.M., and Slifka, M.K. (2003). Nat. Med. 9, 1131–1137.

Miller, J.D., van der Most, R.G., Kondy, R.,Dlidewell, J.T., Albott, S., Masopust, D., Murali-Krishna, K., Mahar, P.L., Edupuganti, S., Lalor,S., et al. (2008). Immunity 28, this issue, 710–722.

Murali-Krishna, K., Altman, J.D., Suresh, M.,Sourdive, D.J.D., Zajac, A.J., Miller, J.D., Slansky,J., and Ahmed, R. (1998). Immunity 8, 177–187.

Seder, R.A., Darrah, P.A., and Roederer, M. (2008).Nat. Rev. Immunol. 8, 247–258.

Terajima, M., Cruz, J., Raines, G., Kilpatrick, E.D.,Kennedy, J.S., Rothman, A.L., and Ennis, F.A.(2003). J. Exp. Med. 197, 927–932.

Immunity

Previews

B Young Again

Thomas Graf1,* and Meinrad Busslinger2,*1Center for Genomic Regulation, Carrer Dr. Aiguader 88, E-08003 Barcelona, Spain2Research Institute of Molecular Pathology, Dr. Bohr-Gasse 7, A-1030 Vienna, Austria*Correspondence: thomas.graf@crg.es (T.G.), busslinger@imp.ac.at (M.B.)DOI 10.1016/j.immuni.2008.04.004

Hanna et al. (2008) report in a recent issue of Cell that a defined set of transcription factors can reprogrammature B cells back to pluripotent stem cells.

The stem cell field was stunned two years

ago when Takahashi and Yamanaka dem-

onstrated that the transcription factors

Oct4, Sox2, Klf4, and c-Myc are capable

of reprogramming fibroblasts into what

was termed induced pluripotent stem

(iPS) cells. iPS cells closely resemble

606 Immunity 28, May 2008 ª2008 Elsevier

embryonic stem (ES) cells in their cellular

phenotype and gene-expression profile

(Takahashi and Yamanaka, 2006) and con-

tribute to the germline when transplanted

into blastocysts (Okita et al., 2007; Wernig

et al., 2007). Because of the low reprog-

ramming efficiency, it remained, however,

Inc.

unclear whether iPS cells originate from

differentiated fibroblasts or rare stem cells

coexisting in the fibroblast culture. A simi-

lar argument was put forward to explain

the low success of previous somatic cell

nuclear-transfer experiments. Hochedlin-

ger and Jaenisch (2002) subsequently

Immunity

Previews

Figure 1. Reprogramming of B Lymphocytes into iPS CellsPrevious experiments demonstrated that viral expression of Oct4, Sox2, Klf4, and c-Myc (4 Yamanaka transcription factors; 4Y TFs) in mouse embryo fibroblasts(MEFs) induces the formation of induced pluripotent stem (iPS) cells (black arrow). It was also shown that forced expression of C/EBPa in mature B cells induces theirtransdifferentiation into macrophagesvia inhibitionof Pax5 function and upregulation of PU.1 expression (red arrow). Furthermore, conditional Pax5 inactivation wasshown to allow mature B cells to dedifferentiate into uncommitted lymphoid progenitors (LP, red arrow). The work of Hanna et al. (2008) now demonstrates that the4Y TFs induce iPS cell formation in pro-B cells and pre-B cells, whereas these factors are ineffective in mature B cells. Mature B cells are only induced to develop intoiPS cells upon coexpression, with the 4Y TFs, of either C/EBPa or an shRNA lentivirus interfering with Pax5 expression (shPax5). Hematopoietic stem cells, HSC.

showed, however, that transplantation of

B and T cell nuclei into enucleated oocytes

yielded blastocysts from which ES cells

could be derived. These ES cells gener-

ated mice in which all cells contained

either immunoglobulinor T cell receptor re-

arrangements, showing that differentiated

B and T cells can be reprogrammed to plu-

ripotency by nuclear-transfer experiments

(Hochedlinger and Jaenisch, 2002). This

study also raised the possibility that mature

lymphoid cells can be reprogrammed by

forced expression of the 4 Yamanaka

(4Y) transcription factors.

This question was addressed in the new

work from the Jaenisch lab, who first stud-

ied the plasticity of immature B cells from

the bone marrow and then that of mature

B cells from the spleen (Hanna et al.,

2008). In brief, the strategy adopted took

advantage of their earlier study, which

showed that fibroblasts from a transgenic

mouse expressing the tetracycline trans-

activator could be reprogrammed into

iPS cells when infected with tetracycline-

inducible lentiviral vectors of the 4Y factors

and treated with doxycycline. Importantly,

these iPS cells contributed to the forma-

tion of chimeras after injection into blasto-

cysts (Brambrink et al., 2008). In the new

work, Hanna et al. (2008) used the resulting

chimeras as a source of pro-B and pre-B

cells, which in each case constitute a mix-

ture of wild-type B lymphocytes and iPS-

cell-derived B cells containing the four len-

tiviral transgenes. These immature B cells

were cultured on OP9 stromal cells in the

presence of lymphoid cytokines to main-

tain B cell growth and, when stimulated

with doxycyclin, developed colonies of ad-

herent cells that became positive for the

ES cell markers alkaline phosphatase and

Nanog (Figure 1). Four of the seven pro-

B-cell-derived iPS cell lines analyzed (iB-

iPS) carried DH-JH rearrangements at the

immunoglobulin heavy chain (IgH) locus.

The presence of a DH-JH rearrangement

does, however, not unequivocally identify

the cell of origin as a pro-B cell because

uncommitted pre-pro-B cells and NK cell

precursors also carry DH-JH rearrange-

ments and are also present within the

B220+c-Kit+ gate used to FACS-sort the

pro-B cells. More convincingly, the pre-

B-cell-derived iB-iPS cell lines exhibited

VH-DJH rearrangements, which were gen-

erated in committed pro-B cells. The iB-

iPS cell lines tested were clearly pluripo-

tent because they induced teratomas

and, upon blastocyst injection, contrib-

uted to the formation of germlinechimeras.

Biopsies obtained from various tissues of

a chimeric mouse exhibited the same Igh

rearrangements as the iB-iPS cell line

used for producing it. These results there-

fore show that the 4Y factors are sufficient

to reprogram B cell precursors into plurip-

otent cells closely resembling ES cells.

Imm

In contrast to the results obtained with

early B cells, mature B cells from the

spleen of adult chimeric mice could not

be reprogrammed with the 4Y factors. In

attempts to overcome this hurdle, the

authorscoexpressedC/EBPa, a transcrip-

tion factor that had been shown to repro-

gram both immature and mature B cells

from the spleen into cells resembling mac-

rophages (Xie et al., 2004; Figure 1). In-

deed, when C/EBPa was introduced into

either mature splenic B cells from a chime-

ric animal containing inducible transgenes

of the 4Y factors, or when mature B cells

were directly coinfected with lentiviruses

of all five factors, B-iPS cell colonies could

be generated (Figure 1). As before, these

B-iPS cells were capable of generating

coat-color chimeras and contributed to

the germline. Each of the nine lines tested

carried a functional rearrangement at the

Igh locus and at one of the genes encoding

Igk or Igl light chain, as it is characteristic

of mature B cells. Strikingly, one line even

exhibited somatic hypermutations in the V

regions of their heavy and light chains;

such hypermutations are a hallmark of

germinal center B cells that have under-

gone immunoglobulin affinity maturation

in response to antigen encounter.

Why is C/EBPa so effective in facilitating

the reprogramming of mature B cells? One

possibility is that the necessary key event

to render mature cells susceptible to

unity 28, May 2008 ª2008 Elsevier Inc. 607

Immunity

Previews

reprogramming is the inhibition of the B

cell commitment factor Pax5. Thus, C/

EBPa is known to antagonize the function

of Pax5 during induced transdifferentia-

tion (Xie et al., 2004). In addition, condi-

tional Pax5 inactivation has been shown

to allow mature B cells to dedifferentiate

in vivo into early uncommitted progenitors

with myeloid and T lymphoid differentia-

tion potential (Cobaleda et al., 2007; Fig-

ure 1). This hypothesis is supported by

the finding that an shRNA lentivirus inter-

fering with Pax5 expression was capable

of replacing C/EBPa in the reprogramming

experiments (Hanna et al., 2008; Figure 1).

Hence, the Pax5-dependent transcription

program, which maintains the identity of B

cells, needs to be disrupted by ectopic C/

EBPa expression or specific Pax5 knock-

down to allow the reprogramming of ma-

ture B cells into iPS cells by the 4Y factors.

The puzzling finding that the 4Y factors

are sufficient to convert pro-B and pre-B

but not mature B cells into iPS cells implies

that one (or more) of the 4Y factors must

be capable of inactivating Pax5 in early B

lymphocytes. In contrast, C/EBPa can re-

program both early and late B cells (Xie

et al., 2004), thus pointing to different

mechanisms by which the 4Y factor(s)

and C/EBPa inactivate Pax5 function. It

is thus conceivable that the Pax5 tran-

scription complex or hypothetical regula-

tory element(s) of the Pax5 gene escape

the inhibitory action of the 4Y factor(s) in

mature B cells. Because 4Y factor(s)

have recently been shown to also repro-

gram other differentiated cell types such

as hepatocytes and epithelial cells of the

608 Immunity 28, May 2008 ª2008 Elsevier I

stomach, it is possible that they impose

an ES cell phenotype by a more general

mechanism (Aoi et al., 2008). How this

works would be interesting to know.

Another question raised by the study of

Hanna et al. (2008) is how Pax5 downregu-

lation renders mature B cells susceptible to

reprogramming. Is it because the cells are

induced by C/EBPa expression to tran-

siently develop into macrophages, which

might be more susceptible to reprogram-

ming because of their adherence? Or, do

mature B cells dedifferentiate to pro-B

and pre-B cells or even further to uncom-

mitted lymphoid progenitors in response

to Pax5 loss? The latter possibility seems

plausible; multipotent progenitors may be

more susceptible to reprogramming be-

cause their chromatin structure and epige-

netic state are likely to be more plastic than

those of terminally differentiated cells.

However, the study of Hanna et al. (2008)

also shows that differentiated mature B

cells havesurprisinglyhigh reprogramming

efficiencies (�3%). Furthermore, hepato-

cytes and gastric epithelial cells undergo

full reprogramming more readily than fibro-

blasts (Aoi et al., 2008), and the nuclei of

postmitotic granulocytes apparently can

be more efficiently reprogrammed than

those of hematopoietic stem cells after

transplantation into oocytes (Sung et al.,

2006). These counterintuitive observations

are a reminder of how little we understand

about the role of transcription factors in ex-

tinguishing one gene-expression program

while activating another one.

The studies of Hanna et al. (2008) have

established a new principle that might be

nc.

key for learning how to overcome the

resistance of differentiated cells to the re-

programming activities of the 4Y factors.

Instead of looking for an additional re-

programming factor acting more proxi-

mally to ES cells, it may be more advisable

to search for a facilitating transcription

factor among those regulators that func-

tion in the developmental context of the

differentiated cell in question.

REFERENCES

Aoi, T., Yae, K., Nakagawa, M., Ichisaka, T., Okita,K., Takahashi, K., Chiba, T., and Yamanaka, S.(2008). Science, in press. Published onlineFebruary 14, 2008. 10.1126/science.1154884.

Brambrink, T., Foreman, R., Welstead, G.G.,Lengner, C.J., Wernig, M., Suh, H., and Jaenisch,R. (2008). Cell Stem Cell 2, 151–159.

Cobaleda, C., Jochum, W., and Busslinger, M.(2007). Nature 449, 473–477.

Hanna, J., Markoulaki, S., Schorderet, P., Carey,B.W., Beard, C., Wernig, M., Creyghton, M.P.,Steine, E.J., Cassady, J.P., Foreman, R., et al.(2008). Cell 133, 250–264.

Hochedlinger, K., and Jaenisch, R. (2002). Nature415, 1035–1038.

Okita, K., Ichisaka, T., and Yamanaka, S. (2007).Nature 448, 313–317.

Sung, L.Y., Gao, S., Shen, H., Yu, H., Song, Y.,Smith, S.L., Chang, C.C., Inoue, K., Kuo, L., Lian,J., et al. (2006). Nat. Genet. 38, 1323–1328.

Takahashi, K., and Yamanaka, S. (2006). Cell 126,663–676.

Wernig, M., Meissner, A., Foreman, R., Brambrink,T., Ku, M., Hochedlinger, K., Bernstein, B.E., andJaenisch, R. (2007). Nature 448, 318–324.

Xie, H., Ye, M., Feng, R., and Graf, T. (2004). Cell117, 663–676.