Volume 12 Issue 2 - January 1, 2010 PDF
Paxillin Is a Target for Somatic Mutations in Lung Cancer: Implications for Cell Growth and Invasion
Ramasamy Jagadeeswaran1, Ravi Salgia1, and Yi-Ching Wang2,*
1Department of Medicine, Cancer Research Center, University of Chicago
2Department of Pharmacology, College of Medicine, National Cheng Kung University
Cancer Res 2008; 68:132-142.
SCI 7.672、Ranking 8/132 (6%) in “Oncology”
Font Enlarge
Paxillin is a 68-kDa adapter protein with five LD motifs at the N-terminus and four tandem LIM domains towards the C-terminus. Paxillin is phosphorylated on specific tyrosine and serine residues mostly located in the N- terminal half in response to growth factors.1 Phosphorylation of these residues results in the generation of specific SH2 interaction sites. Several studies clearly demonstrate a role for paxillin in cell motility and metastasis. Our lab has previously shown significantly elevated levels of paxillin in various lung cancer tissue samples as compared to adjacent normal lun g tissue.2 It is likely that paxillin could contribute both to the lung cancer cell transformation and its metastasis.

In order to understand the role of paxillin in lung tumorigenesis we have systematically examined paxillin expression in lung tumor samples and also determined their relationship to tumor type, stages, cancer progression and metastasis. Paxillin levels increased from normal to dysplasia and to tumor (Fig. 1A). The level of expression of paxillin increased with increasing stage (Fig. 1B), thus clearly implicating paxillin in higher stage tumors—that reflect metastatic setting.
Fig. 1. (A) Representative photomicrograph of paxillin in tissues from normal (1+), Dysplasia (2+) and primary tumor (3+). (B) Frequency of paxillin expression in samples from different stages of lung cancer patients.

To further study the role of paxillin in lung cancer, we determined the mutations of its gene in 191 primary lung and 151 other non-lung tumor samples and 71 cell lines. Bi-directional DNA sequence analysis was carried out and the observed genetic alterations in paxillin are summarized (Fig. 2). We identified a total of 21 somatic paxillin mutations among the lung cancer tissues and cell lines tested. Interestingly, identified paxillin mutations (19/21) were located as two clusters in the region between LD1 and LD2 domains (amino acid residues Pro30 to Gly139) and the region spanning the LIM domains (Fig. 2A). All sequence alterations in this group were heterozygous in the tumor DNA except two cell lines (SK-LU-1 and H820) and four tumor tissues; in each case, paired adjacent normal lung tissue from the same patient showed wild-type sequence, confirming that the mutations are somatic in origin (Fig. 2B-C). The mutational spectra of paxillin were characterized by a high proportion (94% or 32/34) of C:G > T:A transitions, compatible with the mutagenic effects of tobacco carc inogens.3 It appears that Caucasians, African-Americans, and Taiwanese each show their unique mutational spectrum (Fig. 2B). In addition, we have identified single nucleotide polymorphism for paxillin and differences noticed in the various ethnic groups.
Fig. 2. Novel somatic mutations within the paxillin gene in cancers. (A) Schematic diagram of paxillin showing unique somatic mis-sense mutations identified in cancer tissues and cancer cell lines in the context of the functional domains of paxillin. Arrowheads, location of mis-sense mutations. (B) Frequency of paxillin mutations among lung tumor samples from different ethnic groups (P = 0.017, χ2 test) and histology (P = 0.019, χ2 test). (C) Classes of mutations found in human cancer tissue specimens and cancer cell lines in the paxillin gene.

The cytoskeletal protein paxillin was originally identified in focal adhesions and is a natural substrate for several oncogenic t yrosine kinases,4 thereby raising the possibility that it itself could contribute to carcinogenesis. Using H522 cells that lack paxillin expression, we also tested the effect of ectopically expressed wild type and paxillinA127T on cell viability after serum starvation. The viability of cells expressing wild type paxillin was comparable to that of vector control cells. Proliferation and colony formation of cells expressing paxillinA127T was at a significantly higher rate (Fig. 3).
Fig. 3. Transfected H522 cells with empty EGFP vector, wild-type paxillin construct or A127T mutant paxillin construct were plated and anchorage independent cell growth of paxillin mutant was assayed. The colonies were then stained and counted. The left panel depicts a representative experiment. The results presented in the right panel are expressed as the number of colonies obtained in control, Wt Paxillin and A127T paxillin respectively (average ± S.D. of three independent experiments).

These findings are particularly interesting in light of recent evidence implicating paxillin in cancer aggressiveness and it will be important to asses the relevance of this alteration for metastasis. These findings have important implications for developing biomarkers and new molecular targeted therapeutic approaches for cancer showing such alteration.

  1. P. C. Ma et al., Cancer Res 63, 6272 (Oct 1, 2003).
  2. R. Salgia et al., Oncogene 18, 67 (Jan 7, 1999).
  3. P. Vineis et al., J Natl Cancer Inst 96, 99 (Jan 21, 2004).
  4. M. D. Schaller, J. T. Parsons, Mol Cell Biol 15, 2635 (May, 1995)
< Previous
Next >
Copyright National Cheng Kung University