- Top of page
- Materials and methods
Transferrin receptor (TfR, CD71) is an integral membrane glycoprotein that mediates cellular uptake of iron. In most tissues, TfR expression is correlated positively with proliferation and regulated at the post-transcriptional level. The available data regarding the pattern of TfR gene expression in haematological malignancies are very limited. In the present study, we evaluated TfR gene expression at the molecular level in bone marrow (BM) samples of 44 patients with de novo acute myeloid leukaemia (AML) at diagnosis with BM blasts > 85%. TfR mRNA levels were determined by densitometric analysis of quantitative reverse transcription polymerase chain reaction products corresponding to TfR exons 15–17. Each sample was tested in at least two independent experiments. In 13/44 patients, TfR messages were not detected (this is probably an underestimate as some positive results may be attributed to residual normal erythroid cells present in the samples). In 17/44, TfR mRNA levels were low–intermediate, and were high in the remaining patients (14/44). TfR mRNA positivity was significantly associated with older age. No statistically significant correlations were found either with specific French–American–British (FAB) subtypes or attainment of complete remission, incidence of relapse and survival (after adjusting accordingly for age and FAB subtype). The absence of TfR mRNA transcripts in a significant minority of cases suggests that alternative mechanisms of iron uptake may function in AML blast cells.
Iron is essential for cell proliferation. Control of iron acquisition and utilization must be modulated to meet the varied and diverse demands of the cells (Lash & Saleem, 1995; Lieu et al, 2001). The best known mechanism for the uptake of iron in both normal and neoplastic cells entails binding of serum transferrin (the main carrier protein for iron in serum) to a specific transmembrane glycoprotein receptor (transferrin receptor, TfR) (Testa et al, 1993). However, other mechanisms have also been found to operate in several cell types (Qian & Tang, 1995; Richardson & Ponka, 1997; Lieu et al, 2001).
In most tissues, TfR expression is controlled by iron availability at the post-trascriptional level in a manner resembling feedback inhibition: fewer receptors are expressed when iron is abundant and more receptors are expressed when iron is scarce (Chan et al, 1994). However, in erythroid cells, TfR expression is regulated at the transcriptional level during erythroid differentiation and feedback mechanisms related to iron levels do not seem to be important (Chan et al, 1994). Furthermore, TfR expression is related to the proliferative state of the cells as well as the induction of differentiation (Theil, 1990); thus, the number of TfR (CD71) molecules is larger in cells with a high proliferation rate (Kuhn, 1994).
The available data from previous protein studies on TfR (CD71) expression in haematological malignancies indicate that CD71 may be associated with certain features of the neoplastic cells [e.g. T-cell immunophenotype in acute lymphoblastic leukaemia (Koehler et al, 1993) and non-Hodgkin's lymphomas (Das Gupta & Shah, 1990)] but are inconclusive as regards the possible clinical significance of CD71 expression (Medeiros et al, 1988; Koehler et al, 1993; Bradstock et al, 1994). However, there is no evidence about the pattern of TfR mRNA expression and stability in primary malignant cells. In this study, we analysed TfR mRNA expression in 44 patients with de novo acute myeloid leukaemia (AML) at diagnosis (treated on a single protocol) and investigated possible prognostic implications.
- Top of page
- Materials and methods
CD71 (TfR) has been detected on dividing cells of all haematopoietic cell lines and on malignant myeloid and lymphoid cells (Testa et al, 1993). CD71 is generally expressed more strongly by proliferating cells – albeit with varying levels of expression (10 000–100 000 molecules per cell; Inoue et al, 1993); this could probably be attributed to increased demands for iron (Qian & Tang, 1995). In the largest clinical series so far, CD71 expression was detected using indirect immunofluorescence in only 140/323 (43%) childhood acute lymphoblastic leukaemia (ALL) cases (Koehler et al, 1993). Similar results have also been reported for other malignant lymphoproliferative disorders (Habeshaw et al, 1983; Medeiros et al, 1988; Das Gupta & Shah, 1990) and in AML as well (33/58 cases, 57%; Bradstock et al, 1994). Several explanations may account for the absence of CD71 from many malignant clones, the most obvious being low sensitivity of the applied methods. Alternatively, TfR protein molecules could be unstable or misfolded (Kuznetsov & Nigam, 1998), while TfR mRNA messages might be inefficiently processed or translated.
Evidently, a crucial issue arising from the aforementioned observations concerns the pattern of TfR mRNA expression in malignant haematopoietic cells. To our knowledge, the present analysis of TfR mRNA transcripts in primary AML blasts is the first study to address this issue. Interestingly, 13/44 AML cases included in our study (29·5%) were negative for TfR cDNA sequences. This is certainly an underestimate of the actual incidence, as some weakly positive RT-PCR signals may derive from rare, residual normal myeloid cells (despite the fact that BM samples analysed in the present study had at least 85% infiltration by myeloid blasts; even a modest contamination with erythroid cells may give positive results). On the other hand, the possibility of a false-negative result was effectively ruled out as all samples were tested RT-PCR-positive for the ‘control’ (RARα) transcripts (Lion & Kidd, 1998). Unfortunately, CD71 expression was assessed using flow cytometry only in 6/44 patients of our series, precluding meaningful conclusions on the fate of TfR messages in AML blasts.
Altogether, the above-mentioned observations and our results support the hypothesis that in certain AML cases alternative, TfR-independent, iron carrier-mediated pathways might be involved in the cellular uptake of iron, similar to that already described for various other cell types (intestinal mucosal cells, liver cells and transformed cultured cells) (Seligman et al, 1991; Chan et al, 1992; Qian & Tang, 1995; Richardson & Ponka, 1997; Trinder & Morgan, 1997, Lieu et al, 2001). In this context, a new TfR-like family member (transferrin receptor 2, TfR2) has been cloned recently and proposed to mediate the cellular uptake of iron, via different mechanism(s) and affinity than classic TfR (Kawabata et al, 1999). In particular, expression of TfR2 is not regulated by cellular iron status, probably because its mRNA does not contain iron-responsive elements (Kawabata et al, 2000).
The expression of activation antigens in a malignant population is related to cell proliferation; thus, it is reasonable to speculate that this expression may be correlated with tumour behaviour and treatment outcome. In this context, it has been shown that TfR is overexpressed in adriamycin-resistant K562 human erythroleukaemic cells and HL60 human myeloid cells, irrespective of P-glycoprotein expression (Barabas & Faulk, 1993). However, generally, the pattern and clinical significance of CD71 expression in haematological malignancies has not been established conclusively. In the series of childhood ALL cases mentioned above, CD71 expression appeared to have no prognostic implications (patients were treated on a single protocol) (Koehler et al, 1993). In non-Hodgkin's lymphomas (NHL), some studies have suggested that CD71 is expressed in more aggressive histological subtypes of B-cell NHL, and that it may negatively affect survival (Habeshaw et al, 1983; Medeiros et al, 1988). In contrast, other studies have reached different conclusions (Das Gupta & Shah, 1990). In a series of AML patients reported by the Australian Leukaemia Study Group 1994 (Bradstock et al, 1994), CD71 reactivity was not predictive of survival duration.
In the present study, we sought to evaluate whether TfR mRNA expression in de novo AML might be associated with various patient characteristics and with response to treatment. An absolute prerequisite for reliable molecular analysis at the mRNA level is the homogeneity of cell sample under study. For this reason, our study was confined to the analysis of BM samples from AML cases with BM blasts > 85%. To perform quantitative measurements of TfR mRNA, we developed a quantitative RT-PCR assay and correlated the intensity of expression to various patient variables and the response to treatment. Bearing in mind that the number of cases examined was not very large, the following conclusions could be drawn: (i) TfR mRNA expression was closely related to age, with younger patients more likely to be negative for TfR mRNA transcripts; (ii) TfR mRNA expression was not associated with specific FAB subtypes; and (iii) the prognostic effect of TfR mRNA expression on remission, relapse and overall survival was not found to be significant, after adjusting for age and FAB subtype. In future studies, it would also be useful to investigate possible correlations between TfR mRNA expression and iron status of the patients (this kind of analysis was not feasible in our study because of insufficient data). Nevertheless, it should not be forgotten that in the haematopoietic tissue the level of iron is not a critical determinant of TfR expression.
In conclusion, these data support the notion that high expression of TfR mRNA might simply represent a manifestation of the adverse biological profile of AML in the elderly. In this context, it is not surprising that TfR mRNA expression was not found to be an independent prognostic factor. Finally, our results point to the existence of as yet undefined mechanisms ensuring adequate supply of AML blasts with iron, which operate independently of TfR.