In vivo expansion of the endogenous B-cell compartment stimulated by radiation and serial bone marrow transplantation induces B-cell leukaemia in mice


Dr Tessa Holyoake, Department of Medicine, University of Glasgow, Royal Infirmary, 10 Alexandra Parade, Glasgow G31 2ER, UK. E-mail:


Chronic lymphocytic leukaemia (CLL) is a malignancy of CD5+ B cells. This B-cell lineage is established during ontogeny and replenished by the process of self-renewal. Spontaneous and induced leukaemias that frequently affect this lineage are thought to arise as a result of the frequent cell division required to maintain the population throughout adulthood and in response to repeated exposure to environmental antigens. In a series of bone marrow transplant (BMT) experiments performed in B6D2F1 mice, B-cell leukaemia occurred in recipients of serially transplanted syngeneic bone marrow. This study was therefore designed to determine the frequency and phenotype of the observed leukaemia. Male donor cells were initially transplanted into lethally irradiated female hosts and secondary (2°) BMT was performed at 3 months. At 1, 2, 3 and 16 months following primary (1°) BMT, and when 2° BMT recipients developed leukaemia, animals were sacrificed and their tissues extensively examined. These analyses confirmed a host-derived CD5+ transplantable B-cell leukaemia that was initiated in 50% of 1° BMT recipients. With serial passage, the leukaemia became more aggressive and lost CD5 expression, suggesting transformation to a high-grade leukaemia/lymphoma. This previously unreported observation suggests that the combination of radiation and subsequent serial transplantation induces a proliferative stress to the host B-cell compartment that is causative in leukaemic transformation.

Chronic lymphocytic leukaemia (CLL) is the most common human leukaemia found in the western world. The malignancy arises in B-1a cells that are characteristically CD5+/surface IgM+ (Kantor, 1991). Although CLL may remain stable for many years, the malignant clone displays the tendency to evolve into an aggressive large cell lymphoma (Delsol et al, 1981). The underlying cause of this transformation is currently unknown.

In the mouse, the origin and function of CD5+ B-1a cells remains a controversial issue (recently reviewed by Herzenberg, 2000). Conventional B cells appear to develop late during ontogeny and may be continuously replenished from progenitor cells in normal bone marrow. B-1a cells develop earlier, have a distinct tissue distribution pattern and unique surface phenotype (surface IgM+, IgDlo, B220dull/– and CD5+), and maintain their numbers by self-renewal (Hayakawa et al, 1985).

The majority of spontaneous lymphomas and lymphocytic leukaemias arising in mice appear to be of T-cell origin (Shevach et al, 1972), however, Slavin & Strober (1978) first reported the spontaneous development of B-cell CLL in a 24-month-old control BALB/c mouse. The leukaemia was cloned as the BCL1 cell line and is still used to induce B-cell leukaemia in a variety of experimental models (Weiss et al, 1999). It has since been reported that expanded clones of B-1a cells may be found in senescent (greater than 18 months) normal mice and are universally detectable in young New Zealand Black (NZB) mice that show an age-associated oligoclonal to monoclonal expansion, giving rise eventually to B-cell CLL characterized by unique immunoglobulin gene rearrangements and the tendency to grow readily in irradiated or non-irradiated syngeneic hosts (Stall et al, 1988). Although CLL is a disease of the elderly in humans, such clonal expansions of CD5+ B cells are not a common feature and, in general, these populations are polyclonal (Geiger et al, 2000).

Other forms of murine leukaemia may be induced by radiation or benzene exposure or by specific viruses. These include radiation-induced acute myeloid leukaemia (AML), benzene-induced lymphoblastic leukaemias, lymphomas and mixed-lineage leukaemias, and v-erb-B virus-induced pre-B-cell lymphoid leukaemia (Coggin et al, 1988; Miller & Symonds, 1993; Fennelly et al, 1995; Rithidech et al, 1999). For radiation-induced AML, following serial passage into syngeneic recipients, the leukaemic cells grow faster and fewer are required to achieve induction of leukaemia in recipients (Hepburn et al, 1987).

In the mouse, the principle of stem cell self-renewal was established by serial bone marrow transplantation (BMT) (Wu et al, 1968) and, more recently, refined by competitive repopulation assays (Harrison et al, 1978; Harrison, 1980; Harrison & Astle, 1982; Jones et al, 1989; Keller & Snodgrass, 1990; Fraser et al, 1992; Iscove & Nawa, 1997).

Over several years, in serial BMT experiments designed to assess the self-renewal capacity of murine haemopoietic stem cells, we observed a high incidence of B-cell leukaemia following sequential cell passage. This study therefore aimed to determine the frequency and phenotype of the observed leukaemia. We report the development of host-derived B-cell leukaemia in transplanted mice after sequential rounds of transplantation. This novel observation suggests that the combination of radiation exposure and expansion of the host B-cell compartment in vivo predisposes to a relatively high incidence of leukaemia and, thus, could be developed as a model of leukaemogenesis.

Materials and methods

Donor bone marrow cells The donor cells used for the primary (1°) BMT procedure were harvested from the femurs of male B6D2F1 mice aged 6–8 weeks of age. Although 1° recipient mice received either fresh cells or cells cultured ex vivo, the endpoint of interest (i.e. the development of leukaemia) was found not to differ between the groups and therefore, for the purposes of this report, the data derived from individual groups has been pooled.

Serial BMT procedure The serial BMT protocol is shown schematically in Fig 1. The experimental arm comprised 120 female recipient B6D2F1 mice aged 6–8 weeks of age. The control arm comprised 20 sex- and age-matched, non-irradiated, non-transplanted control mice. For 1° and secondary (2°) transplants, female B6D2F1 recipient mice received a total of 12·25 Gy in two fractions, 3 h apart, at a dose rate of 1·13 Gy/min using a Cobalt-60 source. Bone marrow infusions were given as a single injection (0·1 ml) via the tail vein approximately 1 h after irradiation. For 1° BMT the cell dose transplanted was 5 × 105. Three months following 1° BMT, 5 × 106 bone marrow cells harvested from 40 individual 1° mice were transplanted into a further five individual 2° mice (i.e. total of 200 2° BMT recipients). For tertiary (3°) BMT, the mice received only 2 Gy (to ensure recovery of endogenous haemopoiesis in the absence of development of leukaemia) and transplants were performed by intraperitoneal (IP) innoculation of 5 × 106 spleen cells. Although 27 of the 2° BMT recipients developed leukaemia, only 18 of these were each used to transplant a further five recipient mice, giving a total of 90 3° BMT recipients.

Figure 1.

Protocol design for the serial BMT procedure. In the experimental arm, 120 female mice were each transplanted with 5 × 105 male BM cells by intravenous (IV) injection. These mice were designated as 1° BMT recipients. In the control arm, 20 sex- and age-matched control mice were included. These controls were neither irradiated nor transplanted. At 1, 2, 3 and 16 month time points, 12 experimental mice and three controls were sacrificed and analysed. The 2° BMT was performed at 3 months and 40 1° BMT recipients were each used to transplant a further 5 2° BMT recipients using a cell dose of 5 × 106 BM cells. These 2° BMT recipients were then observed and sacrificied only if they became sick with signs of leukaemia. Of 27 sick mice, 18 were used in a 3° transplant procedure in which 5 × 106 spleen cells were injected by intraperitoneal (IP) injection into a further five mice which had received only 2 Gy of irradiation. These 3° recipients were then observed to d 150 or death.

Viral screening of the mouse colony For the duration of these experiments, the mouse colony was screened for the following potential infections: minute virus of mice (MVM), mouse hepatitis virus (MHV), pneumonia of mice (PVM), Sendai virus, Theilers (GDV11), Mycoplasma pulmonis, mouse rotavirus (EDIM) and reovirus type-3 (REO-3).

Analysis of animals suspected to have developed leukaemia Mice that appeared sick following 2° or 3° BMT or that developed obvious splenomegaly were sacrificed. Blood was taken by cardiac puncture and a full blood count (FBC) was performed using an automated cell counter. Peritoneal lavage was then performed to provide peritoneal cells for assessment of immunoglobulin heavy chain rearrangements. Spleen tissue was used for immunophenotyping, fluorescence in situ hybridization (FISH), pathology, assessment for donor-derived haemopoiesis and immunoglobulin heavy chain rearrangements, and to provide cells for 3° transfer. Bone marrow was harvested to allow assessment of donor-derived haemopoiesis and immunoglobulin rearrangements. In selected animals, further tissues including lung, kidney, liver and gut were examined for pathology.

Assessment for donor-derived haemopoiesis and immunoglobulin heavy chain rearrangements The percentage of male donor haemopoiesis and the presence of immunoglobulin rearrangements were assessed by Southern blot analysis using a Y chromosome-specific probe, pY353/B (Bishop et al, 1985), and a probe for J3–4 (Ehlich et al, 1993) respectively. Equal amounts of DNA from the tissues of individual mice were digested with EcoR1 and then separated on 0·8% agarose gels and transferred to Hybond N membrane (Amersham Life Science, UK) according to established Southern blotting techniques. The probes were labelled with 32P using a random primer labelling kit (Pharmacia P-L Biochemicals, USA). Prehybridization for 5 h and hybridization for 16 h were at 42°C in prehybridization buffer [6× saline sodium citrate (SSC), 5× Denhart's (50×, 1% (w/v) bovine serum albumin (BSA)-fraction V, 1% (w/v) polyvinylpyrrolidone, 1% (w/v) Ficoll-400], 0·5% sodium dodecyl sulphate (SDS), 50% formamide, 10 μg/ml sonicated salmon sperm DNA (Sigma Chemicals, Poole, Dorset, UK). Blots were washed twice for 30 min each at 68°C in 2× SSC, 0·1% SDS, followed by washing twice for 30 min each at 68°C in 0·1× SSC, 0·1% SDS. Filters were then exposed to Kodak X-AR5 film for 2–24 h at −70°C. To allow approximate quantification of the proportion of male (donor) to female (recipient) DNA, a standard titration of male DNA diluted into female DNA was included with all test samples. The comparability of loaded DNA was determined by reprobing membranes with a probe for GAPDH.

Immunophenotyping Test cells were first washed in sterile phosphate-buffered saline (PBS) supplemented with 0·1% BSA and 0·1% sodium azide. Cells (106) were then resuspended in 200 μl of PBS/BSA/sodium azide for labelling. Antibody was added at the optimal concentration determined by titration. After 30 min incubation on ice, the cells were washed twice in PBS/BSA/sodium azide, then resuspended in 1% paraformaldehyde. Flow cytometric analyses were performed on a FACScan (Becton Dickinson (BD; San Jose, CA, USA). The antibodies used were purified rat anti-mouse fluorescein isothiocyanate (FITC)-conjugated CD5 (Sigma Chemical) and sheep anti-mouse phycoerythrin (PE)-conjugated IgM (The Binding Site, University of Birmingham, UK).

FISH FISH was performed using the Y-specific JB16/3 paint and the female-specific probe DXWas70 according to the methods described by Fennelly et al (1996).


In earlier studies performed in our laboratories, which involved serial transfer of pooled donor bone marrow between 1°, 2° and 3° recipients, a high incidence of B-cell leukaemia was observed following a third serial transplant procedure. Owing to pooling of bone marrow harvested from several mice prior to each serial transfer, it was impossible to establish an accurate frequency of leukaemic transformation. Therefore, in this study, the serial BMT protocol was designed without marrow pooling to enable an in-depth analysis of the frequency and nature of the leukaemia (see Fig 1). The mouse colony used for these experiments was screened regularly for evidence of virus infection and was persistently negative over the time these experiments were performed.

At 1, 2, 3 and 16 months after the 1° BMT, 12 randomly selected animals from the experimental arm and three from the control arm were sacrificed and analysed in detail. In all animals tested, FBCs were normal and haemopoiesis in bone marrow and spleen was found to be 100% donor (male)-derived. In the control arm, no immunoglobulin heavy chain rearrangements were detected in bone marrow, spleen or peritoneal cells at any of the time points, however, in 4 out of 12 (33%) apparently healthy 1° BMT recipients analysed 16 months following transplantation, immunoglobulin rearrangements were detectable by Southern blot analysis.

At 3 months following 1° BMT, 40 of the 1° animals were sacrificed and their bone marrow used in 2° BMT (Fig 1). Following transplantation, the 2° recipients were observed for signs of leukaemia developing. When animals became unwell they were sacrificed and analyses of different tissues performed. To confirm that the leukaemia was transplantable, groups of five sublethally irradiated female animals (2 Gy) were then innoculated IP with spleen cells from each mouse thought to have leukaemia.

In this experiment, the leukaemic process was initiated in 20 out of 40 (50%) 1° BMT recipients, giving rise, between 160 and 370 d after transplantation, to 27 separate cases of leukaemia within the group of 2° BMT recipients (Table I). Spleen cells were harvested from 18 of these 27 mice and were innoculated into groups of five sublethally irradiated female animals. Within the follow-up period of a further 150 d, samples from 10 out of 18 mice gave rise to transplantable leukaemias that proved lethal between 20 and 95 d (mean 40 d) of IP transfer. It was clear that as the leukaemia progressed it became more aggressive, such that, in the early stages, none of the single groups of five animals innoculated became ill within the follow-up period, whereas all five of the innoculated animals became sick and died within a few days of each other with the most aggressive cases, at around 20 d after transfer.

Table I.  Origin and transfer of leukaemia.
Code for 1° BMT
Code for 2° BMT
Used for 3° BMT
Induction of transplantable leukaemia
within 150 d (Y/N)

For 2° BMT recipients in which leukaemia developed, the percentage of donor (male) cells in bone marrow and spleen fell as the leukaemia progressed, and at the time of sacrifice was 30 ± 8 (mean ± SEM, n = 26)% (Fig 2, lanes 6–10). When spleen cells harvested from leukaemic animals were then used to innoculate sublethally irradiated female mice, a proportion of these animals developed leukaemia and in such leukaemic animals there was virtually no evidence of residual male donor cells in either bone marrow or spleen (Fig 2, lanes 11–15). These findings demonstrated that the transferred leukaemic cells were female and therefore that the leukaemia had been initiated in a previous host's endogenous cells rather than in the original male donor bone marrow. To further confirm the host origin of the leukaemia, FISH was performed using male- and female-specific probes. Figure 3A and B shows representative examples which confirmed that leukaemic metaphases were negative with a Y chromosome-specific probe but positive with the X-specific probe, demonstrating that the leukaemia was female and therefore host derived. Parallel cytospin preparations of these populations were examined morphologically to ensure a homogeneous population of leukaemic cells.

Figure 2.

Assessment of donor origin of the observed leukaemia. Southern blot analysis was performed with the male-specific probe pY353/B. Lanes 1–5 show a titration of male DNA mixed with female DNA. Lanes 6–10 show representative examples from spleens of 2° BMT recipients. Lanes 11–15 show representative examples from spleens of 3° transplants which appeared to have developed leukaemia. The male-specific 1·5 kb band is arrowed and GAPDH loading controls are included for lanes 11–15.

Figure 3.

Fluorescence in situ analysis (FISH) analysis of leukaemic cells. Representative examples of FISH performed on metaphase preparations from spleen cells of a 3° transplant which developed a lethal leukaemia following innoculation with spleen cells from a 2° BMT recipient. Left: FISH performed with the Y-specific JB16/3 paint. Right: FISH performed with the female probe DXWas70. The cells used for FISH analyses were confirmed to be leukaemic by standard morphology.

When leukaemic animals were sacrificed, most had hepatosplenomegaly and all had raised white blood counts (WBC), anaemia and thrombocytopenia compared with age-matched control animals. Pathology revealed widespread infiltration by lymphoid cells in all tissues examined, including peripheral blood, bone marrow, spleen, liver, gut, kidney and lung. In the early stages, the abnormal white cells in peripheral blood and spleen were mature lymphoid cells that co-expressed CD5 and surface IgM (Fig 4), however, as the disease progressed (particularly after innoculation into sublethally irradiated 3° recipients), immature lymphoid blast cells became prominent and these were shown to have lost CD5 expression. The leukaemic blast cells did not express myeloid antigens (Gr-1 or Mac-1), T-cell antigens (CD3) or erythroid antigens (TER-119, data not shown). With progression, the spleens became greatly enlarged and these features suggested that the leukaemia had undergone transformation to an acute form, as happens in patients with CLL. Figure 5A shows the gross pathology of two abnormal spleens compared with a normal spleen and Fig 5B shows the morphology of the leukaemic cells from a spleen tissue touch imprint. Similar morphology was seen for all other tissues examined.

Figure 4.

Phenotype of the observed leukaemia. Representative dotplots of spleen cells from a control (B) and a leukaemic animal (A and C). Dotplot A shows cells labelled with isotype control antibodies and dotplots B and C show cells labelled with anti-CD5-fluorescein isothiocyanate (FITC) and anti-IgM-phycoerythrin (PE).

Figure 5.

Pathology and morphology of the observed leukaemia. (A) Gross appearance of a normal mouse spleen adjacent to two spleens taken from animals with leukaemia. (B) Morphology of the leukaemic cells present in spleen tissue (magnification × 1000).

Having acertained that the leukaemia was consistent with B-cell CLL (i.e. co-expressing CD5 and surface immunoglobulin), the presence or absence of immunoglobulin heavy chain rearrangements was investigated in bone marrow, spleen and peritoneum. For sex- and age-matched controls analysed at 1, 2, 3 and 16 months, only the 6·1 kb germline band was detected (representative example Fig 6). However, in 4 out of 12 randomly selected, apparently healthy, 1° BMT recipients analysed 16 months following transplantation (but not at earlier time points), and in 20 out of 27 (74%) 2° BMT recipients which developed leukaemia, immunoglobulin rearrangements were readily detectable (Fig 6). The pattern of gene rearrangement varied for mice transplanted with cells from different 1° or 2° recipients; however, as expected, for any five 3° transplants developing leukaemia from cells passaged from a single animal, only a single pattern of immunoglobulin rearrangement was observed. Interestingly, in the 1° recipients analysed at 16 months, rearrangements were more prevalent in either spleen or peritoneal cells than in BM, suggesting a similar pattern of disease progression to that seen in New Zealand Black mice (Okada et al, 1991; Marti et al, 1995). The representative example shown in Fig 6 confirms the presence of a second weak band for peritoneal cells (arrowed) that was not detectable by Southern blot analysis for either spleen or bone marrow cells from the same animal. These data confirm that the experimental conditions described here induce clonal B-cell expansion marked by specific immunogobulin rearrangements and with progression to a B-cell leukaemia.

Figure 6.

Southern blot analysis to detect immunoglobulin heavy chain rearrangements. The germline 6·1 kb band is arrowed. Lane 1 shows a single representative example (E) for spleen from a control mouse analysed at 16 months. Lanes 2–4 show results for peritoneal (P), spleen (S) and bone marrow (BM) cells from a single representative 1° BMT recipient sacrificied 16 months after BMT. The arrow indicates an additional weak band present for P (lane 2) but not for S or BM (lanes 3 and 4). Lanes 5–8 show results for spleen samples obtained from individual representative 2° BMT recipients at the time leukaemia developed (A–D).


Following the initial observation that B-cell leukaemia may occur following serial BMT in B6D2F1 mice, this study was designed to determine the frequency and nature of the leukaemic transformation.

The leukaemias arising in 2° BMT recipients were shown to have been initiated in approximately 50% of 1° animals and were associated with immunoglobulin heavy chain rearrangements. It is presumed that irradiation of 1° recipients, coupled with the repeated proliferative stress of repopulation, led to the induction of oligoclonal and then clonal B-cell expansion and later to leukaemia. Animals surviving to more than 1 year following a single BMT procedure showed predominantly (> 95%) donor-derived haemopoiesis in bone marrow, early evidence of immunoglobulin rearrangements in peritoneum and spleen, and, to a lesser extent, in bone marrow, but no evidence of leukaemia. In view of the routine viral screening performed on the mouse colony throughout the time these experiments were undertaken, it is improbable that a virus was responsible for the observed leukaemias.

Murine leukaemias, either arising spontaneously or induced by radiation or drugs, have been described previously and may be strain or age dependent. New Zealand Black mice spontaneously develop a CD5+ B-cell lymphoproliferative disorder as they age and have been used as a murine model of B-cell CLL (Okada et al, 1991; Marti et al, 1995). In these mice, spleen weight, peritoneal cell counts and absolute lymphocyte counts are elevated in old (> 18 months) compared with young (2 months) animals. They develop lymphocytic infiltrates in the lacrymal glands, kidneys, liver and lung. There is an age-dependent increase in CD5+ B cells in blood and spleen which show oligoclonal and even clonal expansion, eventually giving rise to B-CLL (Okada et al, 1991; Marti et al, 1995). To our knowledge, leukaemia of this type has not been described to occur spontaneously in the B6D2F1 strain. Indeed, our control data for mice at ages up to 16 months showed no evidence of immunoglobulin heavy chain rearrangements. Surprisingly, previous reports of serial BMT experiments using the same inbred strain (Jones et al, 1989) have not reported clonal B-cell expansion or B-cell leukaemia. However, there is a single report that transplantation of spleen cells from aged Cba/Ca mice into congenic recipients will induce a leukaemia of this type in 50% of transplanted mice after a long latency period (Fowlis et al, 1989). In our study, 27 out of 200 2° BMT recipients developed leukaemia between 160 and 370 d after BMT. These 27 cases could be traced to 20 out of 40 individual 1° BMT recipients. It is possible that animals included in similar previous studies were not followed for long enough for the leukaemias to compromise their health and, thus, although present, the leukaemia remained undetected.

Mutations probably occur under conditions involving intense and repetitive proliferation when there is less time for proliferating cells to become quiescent and for cell repair mechanisms to come into effect. As the majority of B cells (non-CD5) can be generated from multipotent stem cells present in the bone marrow, regeneration of these cells on transplantation is probably a consequence of proliferation with attendant differentiation from uncommitted progenitors in vivo. In the case of the CD5 lineage, however, the limited ability of stem cells to reconstitute CD5+ B cells upon transplantation would suggest that replenishment of this lineage will be derived from enforced self-renewal of the endogenous CD5+ B cells, previously exposed to radiation, thus increasing the likelihood of a leukaemogenic event. Whether exposure to environmental antigens with which the CD5 B-cell lineage is known to react also contributed to driving development of these clones is unknown.

Controversy still remains regarding the origin and proliferative potential of the CD5+ B-cell lineage that is so frequently represented in spontaneous murine leukaemias and human B-cell CLL. The prevailing view is that CD5+ B cells constitute a separate B-cell lineage that is set up early in development and maintained throughout life by a process of self-renewal. A number of studies have suggested a lack, but not absence, of CD5+ B-cell repopulation potential of transplanted murine bone marrow (Hayakawa et al, 1985; Kantor et al, 1995), indicating that bone marrow is not a major source of the self-renewing CD5 population and also that the long-term engrafting stem cells present in the bone marrow are compromised in their ability to give rise to this lineage upon transplantation.

In summary, we report the interesting observation that serial transplantation of endogenous bone marrow following exposure to radiation, facilitates the outgrowth of a B-cell leukaemia in a significant proportion of BMT recipients. These data may offer the potential to develop a new in vivo model for a common leukaemia in man, B-cell CLL.


This work was supported by grants from the Leukaemia Research Fund (T.L.H.) and the Cancer Research Campaign (Beatson Laboratories). We gratefully acknowledge the excellent technical assistance of Stephen Bell and Tom Hamilton. We should like to thank Dr K. Rajewsky for providing probe J3–4.