Lnk constrains myeloproliferative diseases in mice (original) (raw)
Loss of Lnk enhances and accelerates MPD/CML development in mice. The vast expansion and superior repopulating capability of Lnk–/– HSCs (7, 8) prompted us to study whether Lnk functions as a potential tumor suppressor by limiting HSPC expansion. To test this, we first determined whether loss of Lnk synergizes with oncogenic JAK2, in causing MPDs in mice.
We isolated 5-fluorouracil–enriched (5-FU–enriched) BM cells from both WT and Lnk–/– mice and infected them with either murine stem cell vector–internal ribosomal entry site–GFP (MIG, otherwise known as MSCV-IRES-GFP) vector alone or MIG-TEL/JAK2. When 1 million BM cells were transplanted per lethally irradiated mouse, Lnk–/–;TEL/JAK2 cells elicited an accelerated and exacerbated CML in the host animals, compared with that of WT;TEL/JAK2 cells (Figures 1, 2, 3). While mice transplanted with WT;TEL/JAK2 cells survived well into the second month, mice transplanted with Lnk–/–;TEL/JAK2 cells all succumbed to CML within the first 4 weeks (Figure 1A).
Lnk deficiency exacerbates MPD development initiated by TEL/JAK2 in mice. (A) WT and Lnk–/– BM cells were infected with retroviruses encoding either MIG vector alone or TEL/JAK2, and 1 million total BM cells were transplanted into each irradiated host animal. Kaplan-Meier survival analysis of transplanted mice is shown. P < 0.05, long-rank test comparing WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 groups. n = 5. (B) Lnk deficiency exacerbates CML development initiated by TEL/JAK2 in mice when defined numbers of purified progenitors were transplanted into the hosts. Magnetic bead–enriched Lin– progenitor cells from WT and Lnk–/– mice were infected with retroviruses, and equal numbers of WT and Lnk–/– progenitors were transplanted into irradiated host animals. Kaplan-Meier survival analysis of 3 transplants with different progenitor numbers (left, 50,000 progenitor cells; middle, 30,000 progenitor cells; right, 10,000 progenitor cells) is shown. n = 5 in each group of each transplant.
Lnk deficiency leads to an exacerbated neutrophilia and neutrophil infiltration into multiple organs. (A) The box plot shows wbc measured 2–3 weeks after the transplant (left) and in moribund animals (right). The top and bottom ends of the boxes define the 75th and 25th percentiles, the horizontal lines indicate the medians, and the error bars define the 5th and 95th percentiles. n = 5. P < 0.01 comparing WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 groups. (B) Wright-Giemsa staining of blood smears from WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 mice shows massively increased numbers of myeloid cells at 2–3 weeks. Original magnification, ×400. (C) Spleen weight of transplanted mice (average ± SEM). #P = NS; *P < 0.005. n = 4–6. (D) H&E staining of tissue sections. Original magnification, ×100 (top row, spleen); ×200 (bottom row, liver).
Lnk deficiency leads to an exacerbated neutrophilia, with an expansion of immature myeloid cells in the BM and spleen. Peripheral blood and tissue lineage chimerism analysis after transplantation was performed using flow cytometry. (A) Myeloid chimerism of Lnk–/–;TEL/JAK2 (left) and WT;TEL/JAK2 (middle) mice 2 weeks after BMT and moribund WT;TEL/JAK2 mice (right). Percentages (average ± SEM) of GFP+ myeloid cells are shown. P < 0.001 comparing WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 groups. (B) Lymphoid chimerism of Lnk–/–;TEL/JAK2 and WT;TEL/JAK2 mice. (C and D) Blood lineage chimerism of WT;vector and Lnk–/–;vector mice 3–4 weeks after BMT. Representative percentages of lineage distributions versus GFP expression are indicated. (E–G) Myeloid composition in the BM (E and F) and spleen (G) of WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 mice. (E) The percentage of GFP+Gr-1hi and GFP+Gr-1lo BM cells is quantified (average ± SEM). n = 8. P < 0.01. (F) Mac-1 and Gr-1 expression of GFP+-gated BM cells. The percentage of Mac-1+Gr-1hi mature and Mac-1+Gr-1lo immature myeloid cells is quantified (average ± SEM). n = 8. P < 0.01. (G) The percentage of GFP+Gr-1lo splenic cells (average ± SEM). n = 6. P < 0.05.
Since Lnk–/– mice have 2-fold elevated CFU-GM progenitors in the BM (ref. 6 and unpublished observations), we examined whether the enhanced CML in mice transplanted with Lnk–/–;TEL/JAK2 cells is intrinsic to Lnk-deficient cells or because a greater number of transduced progenitors were transplanted. To examine this, we transplanted equal numbers of purified, infected, lineage negative (Lin–) progenitor cells from WT or Lnk–/– mice into irradiated hosts, along with 200,000 freshly isolated WT competitor BM cells (Figure 1B). Lin– progenitors from both WT and Lnk–/– mice were transduced with similar efficiencies (Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI42032DS1). At each cell dosage, TEL/JAK2-transduced Lnk–/– progenitors conferred reduced survival and accelerated CML compared with TEL/JAK2-transduced WT progenitors (Figure 1B). Thus, Lnk deficiency facilitates MPD development initiated by TEL/JAK2 in mice.
Loss of Lnk leads to exacerbated neutrophilia, with an expansion of the immature myeloid compartment in the BM and spleen. To determine the type of disease that the transplanted mice succumbed to, we examined their hematologic phenotypes. Two to three weeks after transplantation, the wbc counts in mice transplanted with Lnk–/–;TEL/JAK2 BM cells were increased more than 5-fold, compared with mice transplanted with WT;TEL/JAK2 BM cells (Figure 2A, left). This increase was also found in moribund animals (Figure 2A, right). Blood smears confirmed that Lnk–/–;TEL/JAK2 mice exhibited more severe neutrophilia than WT;TEL/JAK2 mice (Figure 2B). Neutrophilia correlated with increased spleen weight (Figure 2C) and neutrophil infiltration into multiple organs (Figure 2D) in both WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 mice, recapitulating major clinical features of CML patients (26). The disruption of splenic structure and increase of spleen weight were greater in Lnk–/–;TEL/JAK2 mice, and neutrophil infiltration into the liver was more diffuse and dramatic (Figure 2D). Therefore, Lnk deficiency exacerbates neutrophilia and neutrophil infiltration into multiple organs.
To extend our morphologic examinations of peripheral blood, we performed flow cytometric analysis, using myeloid and lymphoid markers. We found that almost all GFP+ cells in the blood are Gr-1/Mac-1+ myeloid cells, in both WT;TEL/JAK2 and Lnk–/–;TEL/JAK2 mice (Figure 3, A and B). Importantly, Lnk–/–;TEL/JAK2 mice exhibited a marked expansion of myeloid cells, with over 80% of blood cells being donor derived (GFP+ and Gr-1/Mac-1+); in comparison, fewer than 50% of blood cells were donor derived in WT;TEL/JAK2 mice (Figure 3A). In contrast, mice transplanted with vector-infected WT or Lnk–/– cells showed reconstitution of multiple lineages, including Gr-1/Mac-1+ myeloid, CD4/CD8+ T, and B220+ B cells (Figure 3, C and D).
We next examined lineage reconstitution in the BM and spleens of the transplanted mice. Strikingly, Lnk–/–;TEL/JAK2 BM and spleens exhibited an increased accumulation of Gr-1lo immature myeloid cells in comparison with Gr-1hi mature myeloid cells in WT;TEL/JAK2 mice (Figure 3, E and G). GFP+-infected cells, gated for Mac-1 and Gr-1, revealed an expansion of immature myeloid cells relative to mature ones (Mac-1+Gr-1lo versus Mac-1+Gr-1hi; ref. 27) in Lnk–/–;TEL/JAK2 BM (Figure 3F). This flow cytometric analysis is consistent with BM cytological analyses (Supplemental Figure 2). Thus, Lnk deficiency leads to an exacerbated neutrophilia, with an expansion of immature myeloid cells in the BM and spleen.
Loss of Lnk leads to enhanced progenitor cell expansion in vitro and in vivo and increased cytokine signaling. The marked increase in the number of myeloid cells in Lnk–/–;TEL/JAK2-transduced mice suggests that Lnk deficiency may facilitate myeloid progenitor cell expansion. We thus studied the proliferation and self-renewal of WT and Lnk–/– progenitor cells. Lin– progenitor cells from WT and Lnk–/– BM cells were infected with retroviruses encoding either MIG alone or TEL/JAK2, and purified GFP+Lin– progenitor cells were cultured either in the absence or presence of cytokines (Figure 4). In liquid culture, Lnk–/–;TEL/JAK2 progenitor cells showed enhanced proliferation, both in the absence (Figure 4A) and presence of IL-3 (Figure 4A). When plated in semisolid methylcellulose culture, Lnk–/–;TEL/JAK2 progenitor cells gave rise to similar numbers of colonies as WT;TEL/JAK2 cells in cytokine-free conditions; however, the former generated significantly more colonies in secondary platings (Figure 4B). Vector-infected WT and Lnk–/– progenitors failed to generate any colonies, indicating that Lnk deficiency does not result in cytokine-independent growth. In cytokine-replete conditions, vector- and TEL/JAK2-transduced WT and Lnk–/– cells generated high numbers of colonies in the first plating. However, Lnk–/–;TEL/JAK2 progenitor cells gave rise to significantly more colonies than WT;TEL/JAK2 progenitors in the secondary platings, while vector-infected WT or Lnk–/– progenitors failed to re-plate (Figure 4B). These data suggest that loss of Lnk promotes progenitor cell proliferation and self-renewal in vitro.
Lnk deficiency leads to an enhanced progenitor cell proliferation and self-renewal in vitro. Lin– progenitor cells from WT and Lnk–/– mice were infected with retroviruses encoding either vector alone or TEL/JAK2 and subsequently sorted for GFP positivity. (A) Purified GFP+Lin– progenitors were cultured in the absence of cytokine (left) and in the presence of WEHI supernatant containing IL-3 (right). Live cells were enumerated every 2–3 days, and representative growth curves are shown. The graphs show representatives of 5 independent experiments. (B) Purified GFP+Lin– progenitors were plated in cytokine-free (M3234) or cytokine-rich (M3434) methylcellulose media (StemCell Technologies Inc.). Eight to ten days later, colonies were enumerated, replated, and scored again another 8–10 days later. The total colony numbers of 1st and 2nd platings are shown (average ± SD). *P < 0.01. n = 4–5.
The enhanced proliferation and self-renewal of Lnk–/–;TEL/JAK2 progenitor cells in vitro prompted us to further assess myeloid progenitor expansion in vivo. We isolated BM and spleen cells 2 weeks after transplantation and quantified progenitor numbers in methylcellulose culture assays. We found that TEL/JAK2 expression promoted the expansion of WT progenitors 1.7- and 1.5-fold in the spleen and BM, respectively, when compared with controls. In contrast, TEL/JAK2 caused a 2.5- and 2.6-fold expansion of Lnk–/– progenitors in the spleen and BM, respectively (Figure 5A). Thus, Lnk deficiency predisposed progenitors to TEL/JAK2- induced expansion in vivo.
Lnk deficiency increases the ability of TEL/JAK2 to expand progenitors and enhances cytokine signaling. Two weeks after BMT, spleen and BM cells from mice reconstituted with WT and Lnk–/– cells expressing MIG or TEL/JAK2 (TJ) were isolated. (A) Spleen and BM cells were plated on M3434 methylcellulose media, and CFU progenitors were quantified at days 8–10. Fold differences (average ± SEM) between indicated groups are shown above the bars. *P < 0.01; **P < 0.05. n = 4. (B) BM cells were starved, then either left unstimulated or stimulated with either 10 ng/ml of Tpo or IL-3 for 10 minutes before lysis. Protein lysates were subjected to Western blot analysis. Representative results of 4 independent experiments are shown.
Lnk limits basal level JAK/Stat signaling. We determined whether JAK/Stat signaling was enhanced by isolating BM cells from mice reconstituted with WT or Lnk–/– cells expressing MIG or TEL/JAK2, cytokine starving them, then further incubating them in the absence or presence of Tpo or IL-3 for 10 minutes. We failed to detect signals stimulated with 10 ng/ml Tpo, probably due to limited Mpl expression. Mpl is expressed in LSK cells but only in a small fraction, if at all, of progenitor cells or committed myeloid cells (8, 28). In contrast, 10 ng/ml IL-3 administration robustly stimulated all major signaling pathways (Stat5, Akt, and ERK) in both WT and Lnk–/– groups expressing TEL/JAK2 or vector alone. Importantly, we observed cytokine-independent activation of Stat5 phosphorylation in Lnk–/–;TEL/JAK2 BM cells (Figure 5B). This indicates that loss of Lnk augments constitutive active JAK2/Stat signaling in the transplanted mice expressing TEL/JAK2.
WT Lnk but not SH2-deficient Lnk inhibits MPD/CML development. We previously demonstrated that the Lnk SH2 domain is essential for its inhibitory functions in primary hematopoietic cells (10, 29) and is critical for JAK2 binding (8). In order to dissect the importance of the Lnk SH2 domain in oncogenic JAK2-induced CML/MPD, we performed “rescue” experiments, in which we transduced Lin– progenitor cells from Lnk–/– mice with TEL/JAK2, along with either WT Lnk or a mutant version with an R364E substitution in the SH2 domain (LnkRE). Coexpression of TEL/JAK2 and Lnk was accomplished by using MSCV-TEL/JAK2-IRES-GFP/Lnk expression vectors (Figure 6A). Following transplantation into host animals, we monitored onset and kinetics of MPD development. Lnk–/–;TEL/JAK2-Lnk conferred CML at a much slower rate when compared with Lnk–/–;TEL/JAK2-transduced cells (80 days vs. 30 days) (Figure 6 and Supplemental Figure 3). In contrast, Lnk–/–;TEL/JAK2-LnkRE conferred CML in a similar fashion to that of Lnk–/–;TEL/JAK2 (Figure 6). We conclude that Lnk constrains TEL/JAK2-induced CML in a cell intrinsic manner, at least in part, through its SH2 domain.
WT Lnk but not SH2- deficient Lnk can rescue CML development from Lnk–/– cells expressing TEL/JAK2. (A) Schematic diagram of retroviral constructs used for infection and transplantation. Lin– progenitor cells from Lnk–/– mice were infected with retroviruses encoding either MIG, TEL/JAK2, TEL/JAK2-Lnk, or TEL/JAK2-LnkR364E (TEL/JAK2-LnkRE) and transplanted into irradiated recipients. (B) Myeloid chimerisms of reconstituted mice at indicated time points after transplantation. (C) Survival curves of reconstituted mice.
Lnk deficiency accelerates PV in mice. The V617F mutation in JAK2 has been found at high frequency in MPD patients (16–18). We previously observed that JAK2V617F retains Lnk binding ability, suggesting that Lnk could modulate MPD development resulting from the JAK2V617F mutation. To examine the role of Lnk in MPD development, we isolated BM cells from WT and Lnk–/– mice after 5-FU treatment, infected the cells with either MIG or MIG-JAK2V617F, and subsequently transplanted them into lethally irradiated mice. Mice transplanted with Lnk–/–;JAK2V617F cells developed PV much earlier than those transplanted with WT;JAK2V617F cells (Figure 7A). In recipients of Lnk–/–;JAK2V617F cells, the hematocrits were elevated as early as 3 weeks after transplantation. In contrast, WT;JAK2V617F-transplanted mice developed PV around 5 weeks, as previously reported (30, 31) (Figure 7A). Thus, our data suggest that Lnk deficiency accelerates the onset of JAKV617F-induced PV in mice.
Lnk deficiency accelerates the onset of PV initiated by JAK2V617F and MF progression in mice. (A) The box plot shows hematocrits measured from mice transplanted with WT and Lnk–/– cells infected with retroviruses (either MIG vector or JAK2V617F) at week 3 and week 5. The top and bottom ends of the boxes define the 75th and 25th percentiles, the horizontal lines indicate the medians, and the error bars define the 5th and 95th percentiles. n = 5; 2-tailed Student’s t tests are shown. (B) Lnk restricts myeloid expansion of JAK2V617F-reconstituted mice. WT and Lnk–/– BM cells were infected with JAK2wt, JAK2V617F, or JAK2V617F/Y813F viruses and transplanted into irradiated recipients. Neutrophil counts in the peripheral blood were analyzed 2 months after transplant (average ± SEM). VF, JAK2V617F; VF/YF, JAK2V617F/Y813F. #P = NS; *P < 0.005; **P < 0.05. (C) Reticulin stainings are shown in the BM and spleen sections (top and middle panels), and H&E stainings are shown for the liver sections (lower panel) of transplanted mice. Original magnification, ×200.
Lnk restricts myeloid expansion and MF development in mice. The JAK2V617F construct maintains Lnk binding capability in hematopoietic cells (8), and a physical interaction between Lnk and the JAK2V617F oncogene may directly contribute to restraining MPD in mice. We thus introduced the Y813F mutation into the JAK2VF construct to ablate Lnk binding (Supplemental Figure 4) and determined whether this mutation enhances the ability of JAK2V617F to cause MPD.
Lin– progenitor cells from WT and Lnk–/– mice were infected with retroviruses producing JAK2wt, JAK2V617F, and JAK2V617F/Y813F and transplanted into irradiated recipients. JAK2V617F expression in WT C57/B6 mice primarily confers PV, characterized by moderately increased circulating myeloid and platelet counts (30, 31). In contrast, JAK2V617F expression in Lnk–/– cells triggered a marked expansion of neutrophils in the peripheral blood. Importantly, mice transplanted with WT cells expressing JAK2V617F/Y813F had increased neutrophil counts when compared with WT;JAK2V617F mice (Figure 7B). WT;JAK2Y813F and Lnk–/–;JAK2Y813F mice displayed similar hematocrits and wbc counts to those transplanted with WT or Lnk–/– cells expressing WT JAK2 (data not shown). Together, these data indicate that Lnk deficiency enhances JAKV617F-induced myeloid expansion in mice.
We further examined the progression of the disease into the MF phase. WT;JAK2V617F mice progressed into MF 4 months after transplantation, as previously reported (31). WT;JAK2V617F/Y813F mice progressed into MF (Figure 7C) and anemia (data not shown) faster than WT;JAK2V617F mice, while Lnk–/–;JAK2V617F and Lnk–/–;JAK2V617F/Y813F mice exhibited similar MF status. Specifically, we found that Lnk–/–;JAK2V617F-engrafted mice displayed more severe MF than WT;JAK2V617F-engrafted mice, as evidenced by reticulin staining of the BM and spleen and enhanced myeloid infiltration in the liver at 2 months (Figure 7C). Importantly, WT;JAK2V617F/Y813F mice developed more severe and early onset of MF than WT;JAK2VF mice (Figure 7C). Therefore, our data suggest that Lnk restricts myeloid expansion and the onset of MF in JAK2V617F-reconstituted mice, in part by directly attenuating the JAK2V617F oncogene.
The ability of Lnk to modulate MPD development is not restricted to oncogenic JAK2. Our results above show that Lnk acts in part by directly inhibiting exogenous JAK2V617F oncogene. However, WT cells expressing JAK2V617F/Y813F caused a less severe MPD phenotype than Lnk–/–;JAK2V617F cells, suggesting Lnk might additionally function by inhibiting endogenous signaling pathways in HSPCs. If true, this would imply that Lnk deficiency might cooperate with oncogenes that do not directly rely on JAK2 or bind Lnk. To test this hypothesis, we examined the role of Lnk in modulating CML development caused by BCR/ABL, which does not directly interact with Lnk (32). We found that Lnk–/– BM, transduced with BCR/ABL, resulted in a more aggressive CML when compared with WT BM expressing BCR/ABL (Supplemental Figure 5). Two weeks after transplantation, Lnk deficiency markedly increased neutrophilia and myeloid expansion in the BCR/ABL transplant models (Supplemental Figure 5). Thus, the ability of Lnk to modulate MPD development is not restricted to oncogenic JAK2.
Aged Lnk–/– mice develop a CML-like MPD. Lnk restricts myeloid expansion in transplanted cells expressing diverse oncogenes. This raised the possibility that Lnk–/– mice might spontaneously develop MPDs as they age and accumulate genetic mutations or rearrangements that might promote neoplastic transformation. Young Lnk–/– mice (2 months of age) exhibit a 3-fold increase in platelets as well as wbc (ref. 6 and Figure 8A). This is accompanied by a proportionate increase of both myeloid and lymphoid cells in the peripheral blood, BM, and spleen (6), likely resulting from elevated HSPC numbers. However, old Lnk–/– mice (12–18 months of age) displayed a more dramatic elevation in wbc counts. Moreover, neutrophils and monocytes were expanded disproportionately when compared with other lineages (Figure 8A). Old Lnk–/– mice additionally showed a 5-fold increase in myeloid cells when compared with young Lnk–/– mice and a 9-fold myeloid expansion when compared with age-matched WT mice. In contrast, lymphocyte counts of old Lnk–/– mice were only marginally increased (approximately 1.7-fold), while the platelet counts remained unchanged when compared with young Lnk–/– mice (Figure 8A).
Aged Lnk–/– mice develop a CML-like MPD phenotype. (A) Peripheral blood cell counts of young and old Lnk–/– mice in comparison with age-matched WT mice. Each symbol represents an individual mouse. Average blood counts are indicated by horizontal lines. Young mice are 2 months of age, and old mice are 12–18 months of age. (B) Spleen weight of old WT and Lnk–/– mice. Average spleen weights are indicated by horizontal lines. (C) Percentage of Gr-1+Mac-1+ myeloid cells, CD4+/CD8+ T cells, and B220+ B cells in the spleen of WT and Lnk–/– mice (average ± SEM). n = 5. Two-tailed t tests were performed, and P values are indicated on the plots. (D and E) H&E stainings of the sections from BM located within the femur, spleen, and liver of aged WT and Lnk–/– mice are shown. Original magnification, ×40 (D, center panel); ×200 (E); ×400 (D, left and right panels).
Previous work showed that young Lnk–/– mice exhibit a 2-fold increase in spleen weight, due to an expansion of B cells and megakaryocytes (6, 10). Aged Lnk–/– mice displayed a more pronounced splenomegaly, with a disproportionate expansion of myeloid cells, as determined by flow cytometry and histology (Figure 8, B and C). Specifically, mutant spleens exhibited marked extramedullary hematopoiesis, with disruption of the normal splenic architecture, due to infiltration of mature myeloid cells into the red pulp along with increased numbers of megakaryocytes (Figure 8D). In agreement with these findings, flow cytometry revealed a 5- to 6-fold increase in the percentage of Mac-1+Gr-1+ cells, accompanied by a reduction in B cell percentages. Histological analysis of the BM of aged Lnk–/– mice also revealed increased numbers of granulocytic and monocytic cells (Figure 8D). Finally, periportal cuffing of liver sinusoids with infiltrating myeloid cells was present in the mutant mice (Figure 8E). Taken together, aged Lnk–/– mice developed a CML-like MPD.