Cell cycle-associated accumulation of TIMP-1 in the nuclei of human gingival fibroblasts

目前，我们首次证实了一项早期的免疫组织化学研究，该研究表明TIMP-1样蛋白在人牙龈成纤维细胞(Gin-1细胞)的细胞核中聚集，在细胞周期的S期达到最大(Li等，名古屋医学杂志，58,133-142,1995)。然后，我们从Gin-1细胞的核提取液中分离出该蛋白，经Western blotting、TIMP-1夹心酶免疫分析和基质金属蛋白酶抑制实验证明其与人重组TIMP-1相同。胎牛血清刺激静止期Gin-1细胞48h后，胞液中TIMP-1的含量持续增加，而核提取液中TIMP-1的含量在24 h(S时相)达到最大值，此后显著下降。GIN-1细胞同时表达TIMP-2和TIMP-3以及TIMP-1的mRNA。然而，TIMP-2和TIMP-3蛋白似乎都没有积聚在Gin-1细胞的细胞核中。这些事实很有说服力(…更多的研究表明，TIMP-1以细胞周期依赖的方式在Gin-1细胞的细胞核中特异性积聚。问题是，TIMP-1向细胞核的转运对于不同类型的细胞是否具有普遍性。为了回答这个问题，我们检测了几种人细胞：胎肺成纤维细胞系HFLF、二倍体成纤维细胞系WI-38细胞、骨肉瘤细胞系MG-63细胞和纤维肉瘤细胞系HT1080细胞为成纤维细胞，宫颈癌细胞系HeLa细胞和肝癌细胞系HT1080为成纤维细胞，宫颈癌细胞系HeLa细胞和肝癌细胞系HLE细胞为上皮细胞，观察到TIMP-1蛋白定位于HFLF和WI-38细胞不同步生长的部分细胞核。WI-38细胞在GO期大部分细胞核呈阴性染色。用含10%胎牛血清的培养液替代培养基后，核染色强度随时间而增加，在16h左右达到最大值(可能对应于S时相)。这些观察结果基本上与我们在Gin-1细胞中观察到的结果相同。而在MG-63、HT1080、HeLa和HLE等生长速度较快、细胞周期不清晰的肿瘤细胞中，即使在无FCS的培养液中饥饿48h后，其大部分细胞核仍呈阳性染色。这些事实与公认的概念很好地一致，即许多癌细胞已经失去了将营养有限的细胞送到GO期的控制机制。较少

Presently we first confirmed an earlier immunohistochemical study showing that immunoreactive TIMP-1-like protein accumulated in the nuclei of human gingival fibroblasts (Gin-1 cells), with the maximum in the S phase of the cell cycle (Li et al., Nagoya J.Med.Sci.58,133-142,1995). Then we isolated this protein from a nuclear extract of Gin-1 cells demonstrated it to be identical with human recombinant TIMP-1 by Western blotting, by a sandwich enzyme immunoassay for TIMP-1, and by an assay for matrix metalloproteinase inhibition. The amount of TIMP-1 in the cytosolic fraction after the stimulation by fetal calf serum of quiescent Gin-1 cells increased continuously for 48 hours, whereas, that in the nuclear extract showed a maximum at 24 hours (S phase) and significantly decreased after that. Gin-1 cells expressed mRNAs for both TIMP-2 and TIMP-3 together with that for TIMP-1. However, neither TIMP-2 nor TIMP-3 proteins seemed accumulate in the nuclei of Gin-1 cells. These facts strongly … More suggest that TIMP-1 accumulates specifically in the nuclei of Gin-1 cells in a cell cycle-dependent manner.The question arises as to whether the trafficking of TIMP-1 to the nuclei of cells is general with respect to different cell types. To answer the question, we examined several human cells as follows : fetal lung fibroblasts HFLF,diploid fibroblast cell line WI-38 cells, osteosarcoma cell line MG-63 cells and fibrosarcoma cell line HT1080 cells as fibroblastic cells, and cervix carcinoma cell line HeLa cells and hepatoma cell line HT1080 cells as fibroblastic cells, and cervix carcinoma cell line HeLa cells and hepatoma cell line HLE cells as epithelial cells.We observed that immunoreactive TIMP-1 protein was localized in some of the nuclei of HFLF and WI-38 cells growing nonsynchronously. Most of the nuclei of WI-38 cells showed negative staining at Go phase. After replacing the culture medium with that containing 10% FCS,however, the intensity of the nuclear staining increased time-dependently showing a maximum at around 16h (presumably corresponding to S phase). These observations are essentially the same as we already observed with Gin-1 cells. However, in the sase of tumor cells such as MG-63, HT1080, HeLa and HLE cells which generally proliferate faster than normal cells and have no clear cell cycle, the majority of their nuclei showed positive staining even after starving the cells in FCS-free culture medium for 48h. These facts are in good agreement with the accepted concept that many cancer cells have lost the control mechanism that sends nutritionally limited cells into Go phase. Less