The CLLU1 expression level is a stable and inherent feature of the chronic lymphocytic leukemia clone

Although most leukemias harbor a common genetic aberration, useful for the classification and subgrouping of the disease, no common molecular event leading to chronic lymphocytic leukemia (CLL) has been described. CLLU1 was identified in a comparative screen of patients with and without somatic hypermutations in the variable region of their immunoglobulin receptor heavy chain genes (IgVH). The expression of CLLU1 was shown to be highly upregulated in patients without somatic hypermutations. Subsequent analysis showed that CLLU1 upregulation can be detected only in CLL cells, but not in any other tissue, cell line or in samples from patients with other hematological malignancies.1, 2 In earlier studies we found CLLU1 to have independent prognostic significance in CLL.3, 4 When CLLU1 was analyzed as a continuous variable, a doubling of the CLLU1 expression level was shown to be associated with an increase of 7% in the relative risk of early death.4 Others have since then confirmed the prognostic significance of CLLU1 in CLL.5, 6 The function of the CLLU1 gene is unknown. The gene is not conserved in other species and even though several of the CLLU1 splice variants contain a putative open reading frame of 121 amino acids, we have not yet been able to convincingly detect expression of the putative CLLU1 protein in CLL patient samples. Thus, we do not know whether CLLU1 plays a role in the pathogenesis of CLL, or whether its expression is a reflection of other events in the CLL cells.

The goal of this study was to further investigate the biological properties of the CLLU1 mRNAs. We first developed specific QRT-PCR primer sets to measure the relative expression levels of the different CLLU1 splice variants in samples from 30 CLL patients (Primer and probe sequences are in Supplementary Table 1.) (Figure 1a). The expression pattern of cDNA4/cDNA5 was earlier shown to mirror that of cDNA1,3 and as the cDNA4/cDNA5 Coding Sequence (CDS) PCR reaction does not amplify across an exon–intron boundary, it was not included here. For 25 of the 30 samples we obtained readings in all three QRT-PCR reactions, but in 5 cases we were only able to detect expression of the cDNA1 splice variant. Pair-wise linear regression analysis showed a significant linear relation between the expression levels of all the splice variants (P<0.0001), suggesting that all the splice variants are derived from one primary transcript and regulated by a common promoter upstream from Exon 1 (Figure 1b). As cDNA1 is the most abundantly expressed splice variant (cDNA1>cDNA3>cDNA2/4), we used the cDNA1 QRT-PCR for further analyses. Our results suggest that high-risk CLL patients express all CLLU1 splice variants, whereas low-risk CLL patients may only have detectable expression of the cDNA1 splice variant. The poor prognostic status associated with high CLLU1 expression levels may be a consequence of selective expression of certain CLLU1 splice variants, such as cDNA2/4. Alternatively, the spectrum of outcomes could be the result of a dosage effect incurring a function of the CLLU1 cDNA1 splice variant that is expressed in all CLL samples.

Figure 1
figure1

Expression of the CLLU1 splice variants. (a) Overview of the location of primers used for the CLLU1 splice variant-specific QRT-PCR reactions. (b) The expression patterns of the CLLU1 splice variants were investigated in 25 chronic lymphocytic leukemia (CLL) patients. ΔCt is the number of PCR cycles required to reach the threshold level for CLLU1 after normalization to the β2-M internal standard. ▪—X: cDNA1, Y: cDNA2/4, r=0.90, P<0.0001, Δ—X: cDNA1, Y: cDNA3, r=0.90, P<0.0001, •—X: cDNA2/4, Y: cDNA3, r=0.91, P<0.0001. (c) Blood or marrow samples were collected from the same untreated patient at the same time point (n=15). CLLU1 expression was determined by QRT-PCR (cDNA1 splice variant). Using linear regression analysis we found β=0.89±0.05, r=0.98, P<0.0001. The dotted lines show the 95% confidence intervals.

To determine whether CLLU1 upregulation is a specific feature of CLL cells only, we isolated CD3+ cells (T cells) and CD19+ cells (B cells and CLL cells) from two CLL patient samples, and determined their CLLU1 expression level. In the first sample (UPN100), CLL cells constituted 96% of the CD19+-purified cells and the CLLU1 expression level was 218. T cells constituted 90% of the CD3+-selected sample and this sample had a CLLU1 expression value of 10.3. For UPN 600 the CD19-selected sample had 97% CLL cells and the CLLU1 expression level was 5239. In contrast, the T-cell population constituted 87% of the CD3-selected sample, which had a CLLU1 expression value of 74. These results suggest that expression of CLLU1 is restricted to the CLL cells. We subsequently identified 15 CLL patients, from whom blood and marrow samples were collected on the same date. The samples were Ficoll separated, and the CLLU1 expression levels were determined. Linear regression analysis showed that CLLU1 expression levels were similar in the paired blood and marrow samples (Figure 1c).

The CLLU1 expression level in a sample may be a result of all CLL cells expressing a similar level of CLLU1 or a fraction of cells expressing very high levels of CLLU1. To investigate this we performed in situ hybridization (ISH) on cytospins of Ficoll separated CLL patient samples (n=3). RNA probes were derived by in vitro transcription of a 240-bp fragment of Exon 3 and a 278-bp fragment from the CDS region (Figure 2a). The probes were hybridized to cells from CLL patients with high CLLU1 levels (UPN73 RQ(cDNA1)=1034 (shown in Figures 2b–g), UPN34 RQ(cDNA1)=305 and UPN120 RQ(cDNA1)=324) or to purified normal B-cells. This revealed that the CLL cells had uniform expression of CLLU1, as assessed with either probe (Figures 2b and d). In contrast, we did not see any distinct staining resulting from hybridization of the probes to the control normal B-cells (Figures 2c and e). These results suggest that the measured CLLU1 expression level reflects a common CLLU1 expression level in all the CLL cells within the clone. The in situ hybridization also allowed us to study the cellular localization of the CLLU1 mRNAs within the CLL cells. The cytoplasm of the CLL cells is very small and localized on one side of the nucleus, as illustrated by the Giemsa-stained CLL cells with distinct dark blue nuclear staining (Figure 2g). We detected a distinct cytoplasmic localization of the CLLU1 mRNAs, suggesting that the CLLU1 mRNAs are both spliced and polyadenylated, similar to functional mRNAs (Figure 2f).

Figure 2
figure2

In situ hybridization (ISH) using probes complementary to Exon 3 or CDS sequences of CLLU1. (a) Overview of the CLLU1 gene on 12q22. The two ISH probes are shown in blue. Left panels (b and d) show ISH to sample UPN73, a CLL patient with a CLLU1 expression level of 1034-fold above B-cell levels. The right panels (c and e) show ISH to purified normal B-cell controls. The insets show hybridization with sense probe controls. The bars correspond to 10 μM. (f) Enlargement of ISH-stained chronic lymphocytic leukemia (CLL) cells from UPN73 (exon 3 probe). (g) Giemsa-stained CLL cells.

To address whether the level of CLLU1 changes during the course of the disease, we collected blood samples from untreated CLL patients from at least three consecutive time points. The samples were Ficoll separated and CLLU1 expression levels were determined. Despite the relatively minor fluctuations seen in some samples, all of the patients had low, intermediate or high CLLU1 levels at all the time points investigated. There were some notable exceptions. One patient presented with one clone expressing mutated IgVH 3–23 (89.6% homology to germ line), another rearrangement being unmutated IgVH 1–3 (100% homology to germ line) (closed circles and the dotted line in Figure 3a). At the first time point, at which the CLLU1 expression level was 204, the IgVH 3–23 clone was the dominant band in IgVH analysis. At the later time points the CLLU1 expression level increased to 1397, and this coincided with the VH 1–3 band becoming the strongest band in IgVH analysis, which may be evidence for the presence of two CLL clones in this patient. These results indicate that the CLLU1 expression level of any given CLL clone is stable over time.

Figure 3
figure3

CLLU1 expression is a stable feature of the particular chronic lymphocytic leukemia (CLL) clone. (a) In all, 21 untreated CLL patients were followed longitudinally to determine whether CLLU1 expression changed over time. RQ(cDNA1) indicates the CLLU1 expression level (cDNA1 splice variant) relative to the normal B-pool calibrator. The biclonal patient is marked with closed circles and a dotted line. (b) Cell samples were obtained from 11 CLL patients before treatment and after relapse. White blood cell (WBC) counts in the patient before treatment (pre-Tx), after treatment (post-Tx) and after relapse (post-relapse). (c) CLLU1 expression RQ(cDNA1) measured in CLL samples pre-Tx and post-relapse.

To investigate whether the CLLU1 expression level is affected by treatment, we identified patients who obtained at least a partial response after treatment, with a decrease in the white blood cell (WBC) count to 11 × 109l−1, but then later had relapsed with a considerable increase in the WBC count (Figure 3b). Patients had received fludarabine, fludarabine and rituximab, or steroids and rituximab, and in some cases Campath maintenance. FACS analysis showed that in all the samples CLL cells constituted at least 84% of the cell population (results not shown). Samples from before treatment and after relapse were analyzed for CLLU1 expression, and we found that the CLLU1 expression levels in these samples were similar (Figure 3c). These results show that the CLLU1 level is an inherent marker of the particular CLL clone, and it is not affected by treatment.

The CLLU1 assay was initially developed as an RNA-based one-step QRT-PCR analysis. However, a comparative study of 12 CLL patient samples showed that cDNA synthesis followed by QRT-PCR works equally well (Supplementary Figure 1A). We next collected five tubes of blood from 11 CLL patients, and purified and analyzed the samples successively on days 0–4. The CLLU1 levels were found to be essentially constant over the 5-day period (Supplementary Figure 1B). This suggests that CLLU1 analysis by QRT-PCR is a robust technique that can be successfully applied under circumstances in which transportation or storage of samples over several days is required.

In this report, we show that the expression level of CLLU1 is stable in any given CLL clone. First, we find similar levels of CLLU1 in CLL samples from blood and marrow samples from the same patient. Second, the in situ hybridization experiments show that all the CLL cells within the clone express CLLU1. Third, in our longitudinal study of CLLU1 expression in untreated patients we find that CLLU1 expression is stable over time. Fourth, CLLU1 expression levels are similar in patient samples before treatment and after relapse; hence a high CLLU1 level does not appear to be a feature that is acquired by the CLL clone during the course of the disease. These results show that the level of CLLU1 is a constant and inherent parameter for the CLL clone. As of today, only two molecular CLL clone-specific features are known: the IgVH structure and the CLLU1 expression level. These two parameters may be considered the fingerprint of a particular CLL clone. The biological properties associated with different IgVH structures are currently being unraveled.7, 8 Future research should help elucidate the potential biological implications of varying CLLU1 expression levels.

References

  1. 1

    Buhl AM, Jurlander J, Jorgensen FS, Ottesen AM, Cowland JB, Gjerdrum LM et al. Identification of a gene on chromosome 12q22 uniquely overexpressed in chronic lymphocytic leukemia. Blood 2006; 107: 2904–2911.

  2. 2

    Oppliger Leibundgut E, Dissler D, de Beer D, Marti D, Schuller JC, Duerig J et al. CLLU1 expression distinguishes B-cell chronic lymphocytic leukemia from other B-cell lymphoproliferative disorders. Haematologica 2007; 92 (Suppl 1): 0522.

  3. 3

    Buhl AM, Jurlander J, Geisler CH, Pedersen LB, Andersen MK, Josefsson P et al. CLLU1 expression levels predict time to initiation of therapy and overall survival in chronic lymphocytic leukemia. Eur J Haematol 2006; 76: 455–464.

  4. 4

    Josefsson P, Geisler CH, Leffers H, Petersen JH, Andersen MK, Jurlander J et al. CLLU1 expression analysis adds prognostic information to risk prediction in chronic lymphocytic leukemia. Blood 2007; 109: 4973–4979.

  5. 5

    Hayette S, Hayette S, Bastard C, Mozziconacci MJ, Messana C, Charlot C et al. Expression levels of CLLU1 and LPL: new disease-assessment tools in CLL patients in Binet stage A? Haematologica 2007; 92 (Suppl 1): 0908.

  6. 6

    Chen L, Li J, Zheng W, Zhang Y, Wu Y, Li L et al. The prognostic evaluation of CLLU1 expression levels in 50 Chinese patients with chronic lymphocytic leukemia. Leuk Lymphoma 2007; 48: 1785–1792.

  7. 7

    Stamatopoulos K, Belessi C, Moreno C, Boudjograh M, Guida G, Smilevska T et al. Over 20% of patients with chronic lymphocytic leukemia carry stereotyped receptors: pathogenetic implications and clinical correlations. Blood 2007; 109: 259–270.

  8. 8

    Ghia EM, Jain S, Widhopf II GF, Rassenti LZ, Keating MJ, Wierda WG et al. Use of IGHV3-21 in chronic lymphocytic leukemia is associated with high-risk disease and reflects antigen-driven, post-germinal center leukemogenic selection. Blood 2008; 111: 5101–5108.

Download references

Acknowledgements

This work was supported by grants from The Danish Cancer Society, The Danish Medical Research Council, The Alfred Benzon Foundation, Svend Andersen's Foundation, The Anders Hasselbalch Foundation, The Valdemar and Thyra Foersom Foundation and NIH grant PO1-CA081534. We are grateful to Brian V Hansen, Lone Bach, Trine C Larsen, Sabina Soultanova and Pernille T Sorensen for their technical assistance, and to Monica Cook and Danelle James for their assistance with database searches.

Author information

Correspondence to A M Buhl.

Additional information

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Buhl, A., Novotny, G., Josefsson, P. et al. The CLLU1 expression level is a stable and inherent feature of the chronic lymphocytic leukemia clone. Leukemia 23, 1182–1186 (2009). https://doi.org/10.1038/leu.2009.16

Download citation

Further reading