To the Editor:

Section:

ChooseTop of pageTo the Editor: <<AUTHORS' DISCLOSURES OF P...REFERENCES

We read with interest the study by Domìnguez et al¹ aimed to evaluate the expression of TAp73 and ΔTAp73 variants in patients with colon and breast cancer and their possible associations with E2F-1, p53, and K-ras status. The clinical relevance of alterations in these genes was also assessed. In their breast cancer series, TAp73 suppressor and the ΔTAp73 oncogenic isoforms ΔEx2p73 and ΔNp73 were significantly coupregulated (P = .031 and P = .019, respectively). Statistical associations were also observed between wild-type p53 status and overexpression of the ΔNp73 variant (P = .04), overexpression of E2F-1 and some TP73 forms, and upregulation of ΔTAp73 variants (ΔEx2p73, ΔEx2/3p73, and ΔNp73) and some tumor pathologic markers (vascular invasion and hormone receptor status).

The expression levels of TAp73 and ΔTAp73 were determined by quantitative real-time reverse transcriptase polymerase chain reaction. Furthermore, 14 of the 60 breast cancer cases were also analyzed for TAp73 and ΔNp73 protein expression by immunohistochemistry. TAp73 and ΔNp73 were considered positive in tissue samples exhibiting nuclear staining in more than 10% of the epithelial cells. TAp73 and ΔNp73 immunoreactivity was observed in 21% and 30% of the 14 breast cancer cases, respectively. TAp73 and ΔNp73 nuclear protein expression correlated with mRNA quantification in 12 (86%) and 11 (79%) of the 14 breast samples, respectively.

The significant association between expression of wild-type p53 and upregulation of ΔNp73 observed by Domìnguez et al led the authors to postulate that redundant alterations affecting the same pathway may not confer extra growth advantages on tumor cells during carcinogenesis and thereby alleviate the selective pressure on those cells harboring both alterations.

Our experience refers to the immunocytochemical evaluation of TAp73, ΔNp73, p53, and other biologic variables on 73 fine needle aspirates from primary breast cancer patients. Immunocytochemical analyses were performed on cytospin preparations using the same reagent used by Domìnguez et al: a monoclonal TAp73 antibody (clone 5B429; this antibody recognizes the TAp73 isoforms but does not detect the ΔNp73 variant or p53), a monoclonal ΔNp73 antibody (clone 38C674; this antibody does not cross react with any TAp73 isoform or p53), and a revelation system (LSA; DAKO, Glostrup, Denmark) using diaminobenzidine chromogen as substrate. A monoclonal antibody to p53 (D0-7; Dako Corp, Carpinteria, CA) was used. TAp73 and ΔNp73 immunoreactivity was evaluated in a semiquantitative way. Cells were classified as positive for TAp73 and ΔNp73 when nuclei or cytoplasm or both were scored as 2+ or 3+ in more than 10% of tumor cells. Cells were classified as p53 positive when 10% of the nuclei showed specific nuclear staining. Immunostaining of TAp73 protein was detected in 25 (47%) of 53 carcinomas examined. Staining was confined to the cell nucleus in 10 (40%) of 25 cases and to the cytoplasm in 14 (56%) of 25 cases. One case (4%) showed both cytoplasmic and nuclear TAp73 localization. ΔNp73 immunostaining was detected in 30 (45%) of 67 samples. ΔNp73 staining was cytoplasmic in 28 (93%) of 30 cases and nuclear in two cases (Fig 1). NB4 and K562 cell lines were tested by immunocytochemistry as positive controls for TAp73 and ΔNp73 expression. TAp73 and ΔNp73 expression was confirmed on tumor samples and cell lines by Western blot analysis.

Our results suggest that TAp73 localization may be confined to the cytoplasm and to the nucleus of tumor cells, while ΔNp73 is predominantly cytoplasmic. In this regard, we read with interest the study of Inoue et al² in which a nuclear localization signal and a nuclear export signal were identified in p73, suggesting that p73 localization is controlled by both nuclear import and export and that the overall distribution of p73 is likely to result from the balance between these two processes. Proper control of nuclear import and export is therefore an important regulatory determinant of p73. In contrast, immunostaining of p73 protein was reported to be confined only to the cell nucleus in 41% and 32% of cholangiocellular and hepatocellular carcinomas of the liver, respectively.^3,4 In a more recent report on buccal squamous cell carcinomas, only nuclear p73 expression was considered as positive staining.⁵ With regard to the cytoplasmic ΔNp73 localization observed in our series, the same finding was reported by Uramoto et al⁶ that found that positive expression of ΔNp73 was mainly in the cytoplasm of tumor cells in 77 (58.3%) of 132 patients with lung cancer. In a study on human thyroid cancers,⁷ TAp73 and ΔNp73 expression was predominantly nuclear, although in sporadic cases some cytoplasmic staining was detected. Based on these findings, we suggest that both nuclear and cytoplasmic TAp73 and ΔNp73 staining are considered in tissue sample assessment.

In our series, no correlation was found between TP73 isoforms and estrogen and progesterone receptor status, Ki67 growth fraction, and HER-2/neu amplification whereas, like in the series by Domìnguez et al, a significant correlation between TAp73 and ΔNp73 (P < .05) was observed (Table 1).

Positive p53 immunostaining, suggesting TP53 mutations, was observed in the series by Domìnguez et al¹ in 23 (38%) of 60 breast cancer patients and a statistical association was reported between wild-type status and overexpression of the ΔNp73 variant (P = .04). The fact that in our series, conversely to that of Domìnguez et al, we found a significant correlation between p53 positivity and ΔNp73 expression (P < .05; Table 1) suggests that both alteration, TP53 mutation, and upregulation of the TP73 oncogenic isoform ΔNp73, may not be mutually exclusive, and could confer additional growth advantage on cancer cells.

Significantly, ΔNp73 overexpression was observed in NB4 and K562 leukemia cell lines carrying an inactive p53⁸ and ΔNp73 transcript and protein were detectable in different subtypes of primary acute myelogenous leukemia (AML) blasts.^9,10 In AML, TP53 mutations are quite rare (5% to 10%), but murine double minute 2, its principal negative regulator, has been found to be frequently overexpressed in AML and to inactivate p53^11-13 suggesting that both alterations, inactivation of wild-type p53 protein and ΔNp73 expression, can cooperate in the leukemic process.

In addition, a trend but no significance between overexpression of dominant-negative forms of p73 and concomitant wild-type p53 status on a set of 30 tumors was observed.¹⁴

These findings together with our results suggest that more samples in various type of tumors need to be analyzed to fully clarify the relationship between p53 status, p53 activity, and ΔNp73 protein levels. Finally, further functional studies are needed to elucidate the mechanisms that differentially control TA and ΔNp73 localization, activity, protein stability, and p73-protein interactions, and the role that they play in breast carcinogenesis.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Section:

ChooseTop of pageTo the Editor:AUTHORS' DISCLOSURES OF P... <<REFERENCES

The authors indicated no potential conflicts of interest.