Methylation of SHP1 gene and loss of SHP1 protein expression are frequent in systemic anaplastic large cell lymphoma

Anaplastic large cell lymphoma (ALCL) is a type of non-Hodgkin lymphoma of T/null-cell immunophenotype, characterized by anaplastic cytology and CD30 expression; a subset carries chromo- somal abnormalities involving ALK, and these tumors aberrantly express anaplastic lymphoma kinase (ALK) protein.¹  Several cellular signaling pathways are abnormally activated in ALK⁺ ALCL, owing to the tyrosine kinase activity of ALK.²  Signal transducer and activator of transcription-3 (STAT3), one of the cellular signaling molecules known to be oncogenic, is activated by the nucleophosmin (NPM)–ALK fusion product.³  STAT3 is consti- tutively activated in ALCL, in most ALK⁺ tumors, and approxi- mately half of ALK^- tumors.⁴  We also previously showed that Janus kinase-3 (JAK3), one of the normal physiologic activators of STAT3, is activated in ALK⁺ ALCL cell lines and tumors.^5,6 

As reported in Blood by Chim et al,⁷  gene methylation of SHP1 is frequent in multiple myeloma and is associated with STAT3 activation.⁸  Importantly, restoration of Src homology 2 domain–containing protein tyrosine phosphatase (SHP1) expression by 5-azacytidine treatment is associated with down-regulation of the phosphorylated/activated form of STAT3.⁷  Loss of SHP-1 expres- sion, which correlates with SHP1 methylation, also has been observed in peripheral T-cell lymphomas and types of B-cell lymphoma.^9,10 

SHP1 is known as a negative regulator of STATs and JAKs. It is possible that loss of SHP1 may contribute to STAT3 activation in ALCL, similar to the scenario in multiple myeloma.⁷  We immuno- histochemically surveyed SHP1 expression in ALCL tumors diag- nosed according to criteria defined in the WHO classification.¹  Of 44 cases, 36 ALCLs (16 ALK⁺, 20 ALK^-) (82%) were negative for SHP1 (Figure 1A). There were 8 SHP1-positive ALCLs (3 ALK⁺, 5 ALK^-). Using methylation-specific PCR and primer sets de- scribed previously,⁹  we surveyed 2 ALK⁺ ALCL cell lines, Karpas 299 and SU-DHL-1, and 14 ALCL tumors (5 ALK⁺ and 9 ALK^-). Both cell lines and 9 ALCL tumors (3 ALK⁺ and 6 ALK^-) (64%) showed SHP1 methylation (Figure 1B). None of the 9 cases with SHP1 methylation were SHP1-positive by immunohistochemistry. By contrast, 2 (40%) of 5 cases without SHP1 methylation were SHP1-positive (P = .027; Fisher exact test). These results indicate that loss of SHP1 expression is frequent in ALCL, regardless of ALK status, and that SHP1 gene methylation likely contributes to loss of SHP1 expression.

We also correlated SHP1 expression and STAT3 activation, assessed by immunohistochemistry for tyrosine phosphorylated STAT3 (pSTAT3).⁴  Using a 10% cutoff, 23 of 35 SHP1-negative cases and 5 of 7 SHP1-positive cases were pSTAT3-positive. Among pSTAT3-positive cases, the median percentage of pSTAT3-positive tumor cells was 80% in the SHP1-negative group, compared with 50% in the SHP1-positive group (P = .057; t test). These data suggest that loss of SHP1 expression contributes to increased STAT3 activation in ALCL tumors.

In summary, our data from studies of ALCL parallel those of Chim et al,⁷  who studied multiple myeloma. It is likely that lack of SHP1 expression contributes to constitutive activation of the JAK/STAT pathway in other tumor cell types and enhances the oncogenic potential of STAT3. It is of note that 3 cases in our series lacked SHP1 protein expression but had no evidence of SHP1 methylation, suggesting that alternative mechanisms that silence SHP1 expression are also likely to exist.