Senescence-accelerated Mouse (SAM): With Special Reference to Development and Pathobiological Phenotypes

Toshio Takeda, M.D., is Chairman of the Council for SAM Research, c/o Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. Keichi Higuchi, Ph.D., Lecturer, and Masanori Hosokawa, M.D., Associate ProlEssor, are with the Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. Toshio Takeda, M.D., is Chairman of the Council for SAM Research, c/o Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. Keichi Higuchi, Ph.D., Lecturer, and Masanori Hosokawa, M.D., Associate ProlEssor, are with the Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. Toshio Takeda, M.D., is Chairman of the Council for SAM Research, c/o Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. Keichi Higuchi, Ph.D., Lecturer, and Masanori Hosokawa, M.D., Associate ProlEssor, are with the Department of Senescence Biology, Chest Disease Research Institute, Kyoto University. The development of a senescence-accelerated mouse (SAM 1 ) strain has been under way at Kyoto University since 1970. At present, 12 lines of inbred strains have been maintained as original colonies of the SAM model: 9 senescence-prone (SAMP 1 ) and 3 senescence-resistant (SAMR 1 ). In addition, several lines of SAM have been developed at Takeda Chemical Ind., Ltd., using original mice from Kyoto University. Based on data obtained from Kyoto University and elsewhere, this review describes (1) the circumstances related to the development of SAM mice and their genetic background, (2) the pathobiological phenotypes of SAM, and (3) the significance of SAM development. In 1968, several pairs of AKR/J mice were donated by Jackson Laboratory (Bar Harbor, Maine) to the Department of Pathology (currently the Department of Senescence Biology), Chest Disease Research Institute, Kyoto University, Japan. While continuing brother-sister mating to maintain this inbred strain under conventional conditions, we became aware that in certain litters most of the mice showed a moderate to severe degree of decreased activity, hair loss and dull hair, coarse skin, periophthalmic lesions, and increased lordokyphosis. Additionally, the life span for most mice was shortened. We also noted the inheritance of these phenotypes. In 1975, 5 litters of mice with severe exhaustion were selected as the progenitors of the senescence-prone (P) series. Litters in which the aging process was normal were selected as progenitors of the senescence-resistant (R) series. Thereafter, selective breeding based on senescence, life span, and pathological phenotypes was carried out in addition to the routine brother-sister mating ( Hosokawa and others 1997 ; Takeda and others 1981 , 1991 ). From each of the selected litters of mice with severe exhaustion, 5 different series were obtained and designated P-1, P-2, P-3, P-4, and P-5; and from each of the selected litters of mice that demonstrated normal aging, 3 different series were obtained and designated R-1, R-2, and R-3. Data from survival curves, Gompertz equation, and grading scores of senescence, as well as growth patterns of body weight of these P and R series, revealed a characteristic feature of aging common to all the P series-- “accelerated senescence.” After a brief period of normal development, accelerated senescence is an irreversible advance of senescence manifested by gross lesions and such signs as abnormal behavior, various skin lesions, and increased lordokyphosis. The P series was named SAM, and our first paper on the SAM model appeared in Mechanisms of Ageing and Development in 1981 ( Takeda and others 1981 ). At that time, the mean life span in the P series was 9.7 months, 27% shorter than that of the R series (13.3 months). The senescence grading score of the P series at 8 months of age was 3.67, about 3 times that of the R series (1.26). The Gompertz coefficient for the P series was 0.133, significantly greater than that for the R series (0.083). Necropsy findings in both series showed prominent amyloid deposition, abscess, thymic and nonthymic lymphoma, and pneumonia. A low incidence of thymic and nonthymic lymphoma in both the R (16.3%) and the P (8.3%) series in comparison with that reported in the authentic AKR/J strain ( Festing 1992 ) suggested a genetic deviation from the original AKR/J in this model of accelerated senescence. A high incidence of amyloidosis in the P (77.9%) and the R (32.0%) series compared with the reported incidence in the AKR/J strain ( Ebbesen 1974 ) also seemed to support this ( Takeda and others 1981 ). Breeding among P-4, P-5, and R-3 mice has been unsuccessful; however, it has been possible to establish several new strains that fulfill the criteria of inbred strains: successful brother-sister mating over 20 generations with a stable homozygosity and stable expression of pathological pheno-types ( Hosokawa and others 1984 ; Takeda and others 1991 ). After 1984, the established strain was denoted SAM-R/1 instead of R- 1 ( Hosokawa and others 1984 ). In 1991, SAM-R/1 changed to SAMR1, following the rules for international nomenclature of inbred strains of mice ( Festing 1992 , 1993 ). In 1994, we established 2 new inbred strains, SAMR5 and SAMP11, from the P-4 and P-3 series, respectively. At present, there are 12 lines, all of which originated from the P and R series. Senescence-prone (SAMP) inbred strains include SAMP1, SAMP2, SAMP3, SAMP6, SAMP7, SAMP8, SAMP9, SAMP10, and SAMP11, while the senescence-resistant (SAMR) inbred strains are SAMR1, SAMR4, and SAMR5. SAM has become a general term for both the SAMP and the SAMR strains ( Takeda and others 1991 , 1994 ). Recent data show that the median survival time of SAMP mice is 9.7 months, 40% shorter than that of the SAMR strains (16.3 months). The senescence grading score of SAMP strains at 8 months is 7.97, about twice that of SAMR (3.94) ( Takeda and others 1994 ). SAMR5 demonstrated a lower incidence of lymphoid cell neoplasms and histiocytic sarcoma, an advantageous characteristic of a senescence-resistant strain, compared with SAMR1 and SAMR4. SAMP11 has the shortest life span among all the SAM strains. The incidence of hyperplasia of the bile duct increased and the life span shortened in each succeeding generation of SAMR2, a resistant strain that is no longer being bred ( Takeda and others, 1994 ). To establish a specific-pathogen-free SAM strain, breeding pairs of the P and R series (P-1, P-2, R-1) and SAMP strains (SAMP6, SAMP10) were sent from Kyoto University to Takeda Chemical Ind., Osaka, Japan. From these series and strains, SAMP1TA, SAMP6/Ta, SAMP8/Ta, SAMP10//Ta, and SAMR1TA have been developed and maintained. The lineage of each strain of SAM from the original AKR strains is presented in Figure 1 . At present, breeding pairs of most of the strains reared at Takeda Chemical Ind. as well as Kyoto University are available for investigators if application is approved by the Council for SAM Research. Lineage of SAM development. +, unsuccessful breeding or stopping of breeding. Necropsy findings on the P and R series suggested genetic deviations in the SAM model from the original AKR/J strain as noted above ( Takeda and others 1981 ). Data on 17 biochemical and 10 immunogenetic markers revealed that at least 1 genotype in each SAM strain differed from the authentic AKR/J strain ( Takeda and others 1994 ). These results, along with information gathered from genome mapping using oligonucleotide probes designed to recognize endogenous mouse retrovirus sequences, show that each SAM strain is genetically distinct and contains certain amounts of genetic information derived from strains other than AKR/J ( Kitado and others 1994 ). Therefore, it is reasonable to assume that the original AKR/J strain was unintentionally outbred with other strains before 1975, when the progenitors of the P and R series were selected and selective breeding began ( Finch 1994 ; Kitado and others 1994 ; Takeda and others 1991 ). At present, it is not clear what strains were mated with the AKR/J mice. Routine postmortem examinations and a series of systematically designed studies conducted on living as well as dead SAM mice revealed that in addition to accelerated senescence, which is common to all the SAMP strains, SAMP strains manifest various pathobiological phenotypes that are often characteristic enough to differentiate the strains ( Table 1 ) ( Takeda and others 1997 ). Some of these phenotypes are briefly presented below. Pathobiological phenotypes Includes nonthymic lymphoblastic lymphoma, immunoblastic lymphoma, and follicular center cell lymphoma. Pathobiological phenotypes Includes nonthymic lymphoblastic lymphoma, immunoblastic lymphoma, and follicular center cell lymphoma. Severe age-associated systemic amyloidosis was first reported in the P series ( Higuchi and others 1983b ; Takeda and others 1981 ; Takeshita and others 1982 ). A unique senile amyloid fibril protein (AApoAII, formerly called AS sam ), which is distinguishable from murine protein AA in secondary amyloidosis, was isolated from the liver of SAMP1 mice ( Matsumura and others 1982 ) and was later found to be present universally in such aged mice, regardless of strain ( Higuchi and others 1991b ). The protein consists of a single polypeptide chain of 78 amino acid residues, and its amino terminus is blocked with pyrrolidone carboxylic acid. Additional biochemical and immunochemical studies revealed that the primary structure of apolipoprotein A-II (apoA-II) is identical to that of the amyloid fibril protein ( Higuchi and others 1983a , 1986 ; Yonezu and others 1986 ). This finding proves conclusively that apoA-II, one of the major protein constituents of plasma high density lipoprotein, deposits in tissues without degradation. Next, apoA-II cDNA of SAMP1 and SAMR1 were characterized. Nucleotide sequencing confirmed the amino acid substitution (Pro 5 -Gln) in SAMP1--a result of the substitution of 2 nucleotides (CCA to CAG) ( Kunisada and others 1986 ). A series of molecular genetic and biochemical studies indicated that 3 variants of apoA-II protein (types A, B, and C) with different amino acid substitutions at 4 positions are present in SAM strains. SAMP1, SAMP2, SAMP7, SAMP9, SAMP l 0, and SAMP11 strains with a high incidence of senile amyloidosis have type C apoA-II (Apoa2 c) with gluta-mine at position 5. Other strains of SAM with a low incidence of senile amyloidosis have type A (Apoa2 a ) or type B (Apoa2 b ) apoA-II, both of which have proline at position 5 ( Higuchi and others 1991a ). Examination of the genotypes of apoA-lI gene and amyloid deposition in the F1 and F2 strains, and mice backcrossed between strains with Apoa2 c and Apoa2 a or Apoa2 b , showed that (1) murine senile amyloidosis is linked to Apoa2 c and transmitted in an autosomal dominant manner with incomplete penetrance, and (2) amyloidosis is far more severe in mice homozygous for Apoa2 c ( Higuchi and others 1991a ; Naiki and others 1993 ; Yonezu and others 1987 ). A kinetic analysis of amyloid fibrillogenesis in vitro was performed using a novel method based on the unique characteristics of the fluorescent indicator thiofiavine-T ( Naiki and others 1989 ). The results indicated that extension of amyloid fibrils proceeds by consecutive association of precursor proteins that attach onto the ends of existing fibrils ( Naiki and others 1991 ). In addition to senile amyloidosis, secondary amyloidosis with AA amyloid deposition was often observed in SAM mice with inflammatory lesions, particularly in SAMP2 and SAMP6 ( Higuchi and others 1983b ; Takeda and others 1997 ). A series of behavioral studies on SAMP8 and SAMP8/Ta disclosed age-related behavioral disorders, such as deficits in learning and memory (Miyamato and others 1986; Yagi and others 1988 : Ohta and others 1989 ; Flood and Morley 1992 ; Flood and Morley 1993 ), emotional disorder, and abnormality of circadian rhythms ( Miyamoto 1997 ; Miyamoto and others 1992 ). Characteristic emotional disorder in SAMP8/ Ta was assessed by elevated plus-maze test, water-drinking conflict, and neophobia test. lngram and others (1994 ), however, did not find behavioral impairment in the SAMP8/Ta in swim maze performance. Ikegami and others (1992 ) reported no age-accelerated decline in radial-arm maze performance in SAMP8/Ta mice, although they showed a marked age-accelerated deficit in acquisition performance relative to SAMR1TA in a passive avoidance task, suggesting a dissociative effect of aging in spatial learning and passive avoidance performance. Morphological, mostly age-related, changes observed in SAMP8 and SAMP8/Ta were summarized as follows: periodic acid-Schiff-positive granular structures, spongiform degeneration in the brain stem, β/A4 proteinlike immunoreactive granular structures, reduction of dendritic spines of hippocampal pyramidal neurons, spheroid in gracile nuclei, lipopigmentation, thalamic neuronal inclusion, and astrocytosis ( Akiguchi and others 1994 ; Akiyama and others 1986 ; Sugiyama and others 1987 ; Takemura 1993 ; Yagi and others 1989 ). Neurochemical studies on SAMP8 brains performed by Dr. Nomura and his colleagues demonstrated that (1) the contents of glutamate (Glu) and glutamine (Gln) in hippocampus were greater than those of SAMR 1; (2) high K + -evoked release of Glu, aspartate, Gln, and alanine was increased; (3) N-methyl-D-aspartic acid (NMDAa)-induced release of [ 3 H]-acetylcholine and [ 3 was the of several such as 1 and and the of protein C were the of [ 3 was greater than that of and proteins that with the C of amyloid precursor protein were increased with aging ( and others 1992 ; Nomura and others 1994 ; and Nomura ; and others 1992 ). In addition, Flood and others ) observed an age-related reduction in to of and when avoidance in also showed that in the hippocampus in for the age-related decline in memory in SAMP8/Ta mice ( Flood and others ). In addition, a series of studies was performed to the of age-related deficits in learning and memory and other behavioral ( and others ; Flood and others ; and others 1994 ; and others 1994 ; and 1993 ; and others 1992 , ; and others 1994 , ; and others 1994 ; and others 1993 ; and others ). In and SAMP10//Ta, age-related impairment of learning and memory ( Miyamoto 1997 ; and others 1992 , 1993 ), emotional disorder, and behavior, as assessed by and circadian rhythms were observed ( Miyamoto 1997 ). In addition to the behavioral age-related brain was observed in ( and others 1992 , 1994 ) their life with brain by relative to 4 months of revealed that the most to age-related in was the of the which showed relative to 2 months of Other and all were found to be ( and others 1994 ). The mean cell of in and of the with the in the decreased in by relative to 2 months of age the life was no age-related in the of or in the ( and others 1992 ). and others ) reported that there was an age-related decline in growth in the of mice at the age of 10 it became that mice also showed age-related deficits in learning and memory ( and others ; and others ). of the strains with deficits in learning and memory with aging such as or of and they for studies of aging of the particularly in with and they new the of In the of development of SAM, we became aware of in a aged SAM mice and began to systematically the age-related changes in in several strains of these strains showed patterns of that by the of the a when the mice were 4 or 5 months of age and with age to more than 20 SAMP6 a significantly lower than other strains. with a low however, the of in as other strains. and contents weight of showed among the strains of these results, along with the findings of and revealed that was not to to ( and others 1986 ). from the of loss based on observed in SAMP6 be to the senile in with type B as by and ). and others ) demonstrated by a among strains of the at the not at the for mice to Therefore, in between SAMP6, SAMR1, and SAMP2 are at least in by in the The lower be to to the decreased in the of the In and in vitro studies by and others (1994 ) revealed that changes in SAMP6 are to the of and others ) showed that the of progenitors in was from SAMR1TA at 1 of a was found at 3 to 4 months of and that impairment of was with decreased and decreased as by analysis of and respectively. In addition, a series of studies to the of the in SAMP6 has been conducted ( and others 1989 ; and others ; and others ; and others , 1993 ). A of the of the in several strains of SAM revealed that all strains 1 ). the SAMP strains changes and more than the SAMR strains. mice are the first to show changes at to 9 months of age and over 12 months of and have the incidence of severe changes with ( and others 1989 ). of studies designed to the of and on the development of suggested that on the which as a a in the of in SAM in with accelerated senescence, which as a systemic ( and others 1994 ). The of mice began to at 10 of of and of of of the revealed that of these were and that showed a characteristic of the degeneration and of the and of the of the is a not the in the of in mice ( Hosokawa and others 1988 , 1993 ). The of the the of the a of and to increased and in the to the of ( and others 1994 ). and others (1992 ) revealed that showed clear with age in SAMP2 and that an age-related in and an in of the be in the of in SAMP1 and SAMR1 showed an age-related loss as elevated of and and decreased of as assessed by ( and others 1994 ). SAMP1 showed a more and severe loss with age than did studies showed an age-related in cell density as well as in the of in SAMP1 and SAMR1 ( and others 1994 ). age-related loss of the and hair was also observed in both SAMP1 and The changes in the and hair however, appeared and more in SAMP1 than in SAMP1 showed greater age-related of the than did These results that impairment in SAM is to a of and as well as to ( and others ). and studies of the showed senile of the in SAMP1 and however, revealed no of of the or in the The changes in both SAMP1 and SAMR1, in an accelerated manner in and senile of the in the senescence in both strains ( and others 1994 ). and others (1994 ) reported that SAMP2 mice also manifested most of the characteristic changes in senile From a series of in showed that primary and to 1 ) were significantly more by months of age in SAMP1 and SAMP2 than in SAMR1 and strains, such as and B ( and 1988 ). and others ) reported that from SAMP1 mice showed an in 1 ) 2 and an of of and were in the of SAMP1 with The of was not as severe as that observed in the or mice was distinct enough to SAMP1 from revealed age-associated and lesions with granular and deposition in SAMP1 mice. To the of the observed in in vitro were became that SAMP1 mice a in to a at as as 2 months of and a at a later age ( Hosokawa and others ). This was to be related to of while other such as and and cell were also normal ( Hosokawa and others ). A series of studies to the genetic of the low in SAMP mice was also performed in The low was by one of which be the and a on while the other has not been ( and others 1989 ; and others 1991 ; and others 1994 ). and others (1994 ) reported that a to that in SAMP1 was found in SAMP8 mice, which show characteristic deficits in learning and memory with low cell was observed in SAMP8 ( and others 1994 ). it has been reported that the of the dendritic and B is in mice in the in vitro ( and others ). and others ) a review of the in SAM with a of the between aging and the age-related of pathological phenotypes. a granular and were in to the of the were by a of of the SAMP strains as the most finding ( Takeda and others 1997 SAMP1 and Based on the it was suggested that a in the of in Necropsy findings the of SAM mice revealed that and a incidence of thymic lymphoblastic lymphoma than other SAMR1 and showed a high incidence of nonthymic lymphoma nonthymic lymphoblastic lymphoma, immunoblastic lymphoma, and follicular center cell and ( Takeda and others 1997 ). and were the most common in SAMR1 and respectively. Dr. and colleagues have reported age-related such as senile amyloidosis, and in SAMP2 and SAMR1 mice ( and others 1987 , 1994 ). in to a on the of and particularly that have not been This is for a model of aging in which aging and without ( Takeda and others 1981 , 1991 ). we have in SAMP strains with accelerated senescence and age-associated in addition to SAMR strains with normal aging for nonthymic and histiocytic in aged SAMR1 and ( Takeda and others 1997 ). far as we 2 that manifest accelerated senescence have been and , b ) senescence in a of The a senile with of the and skin and lesions such as and ) reported accelerated aging in the B strain of mice, which and has severe weight loss and a shortened life The to in this strain of mice decreased with there is no more age-related information on these strains, the SAM model (SAMP and to be the model available for on both accelerated and normal aging ( Takeda and others 1991 ). The SAM model most criteria for the of for aging such as life life span, conditions, and genetic as by ). The of mice to the of aging and age-associated has the of genetic information available on this and the of mouse and ( 1994 ). of the in which were as a of by and others are in the pathobiological phenotypes in of and The support for our that the SAM model is and for aging The SAM model has for on aging as SAM is with to the that are in the (1) accelerated senescence ( and others 1994 ; and others ; Hosokawa and others 1994 ; and others 1988 ; and others 1994 ; and others ; Nomura and others 1989 ; Yagi and others ; and others ), (2) normal aging, (3) the of age-associated and to or the advance of senescence and in age-associated ( and others ; and 1997 ; and others 1991 ; and others ; and others 1994 ; and others ; Miyamoto and others 1994 ; and others 1994 ; and others ; and others ; and others , ; and others ). This was in by from the of and and the of and Japan. The the of both the Takeda Chemical Ltd., and the Laboratory of for and Chest Disease Research Institute, Kyoto University, for their