Cytokine Gene Expression As a Function of the Clinical Progression of Alzheimer Disease Dementia

ORIGINAL CONTRIBUTION Cytokine Gene Expression as a Function of the Clinical Progression of Alzheimer Disease Dementia

James D. Luterman, PhD; Vahram Haroutunian, PhD; Shrishailam Yemul, PhD; Lap Ho, PhD; Dushyant Purohit, MD; Paul S. Aisen, MD; Richard Mohs, MD; Giulio Maria Pasinetti, MD, PhD

Background: Inflammatory cytokines have been linked gyrus (PϽ.01). When stratified by the Consortium to to Alzheimer disease (AD) neurodegeneration, but little Establish a Registry for Alzheimer’s Disease (CERAD) is known about the temporal control of their expression neuropathological criteria, IL-6 mRNA expression in in relationship to clinical measurements of AD demen- both the entorhinal cortex (PϽ.05) and superior tem- tia progression. poral gyrus (PϽ.01) correlated with the level of neurofibrillary tangles but not neuritic plaques. However, in Design and Main Outcome Measures: We mea- the entorhinal cortex, TGF-␤1 mRNA did not correlate sured inflammatory cytokine messenger RNA (mRNA) with the level of either neurofibrillary tangles or neu- expression in postmortem brain specimens of elderly sub- ritic plaques. Interestingly, in the superior temporal jects at different clinical stages of dementia and neuro- gyrus, TGF-␤1 mRNA expression negatively correlated pathological dysfunction. with neurofibrillary tangles (PϽ.01) and showed no relationship to the pathological features of neuritic Setting and Patients: Postmortem study of nursing plaques. home patients. Conclusions: The data are consistent with the hypoth- Results: In brains of cognitively normal control sub- esis that cytokine expression may differentially contrib- jects, higher interleukin 6 (IL-6) and transforming ute to the vulnerability of independent cortical regions growth factor ␤1 (TGF-␤1) mRNA expression was during the clinical progression of AD and suggest that observed in the entorhinal cortex and superior temporal an inflammatory cytokine response to the pathological gyrus compared with the occipital cortex. Compared effects of AD does not occur until the late stages of the with age-matched controls, subjects with severe/ disease. These findings have implications for the design terminal dementia, but not subjects at earlier disease of anti-inflammatory treatment strategies. stages, had higher IL-6 and TGF-␤1 mRNA expression in the entorhinal cortex (PϽ.01) and superior temporal Arch Neurol. 2000;57:1153-1160

EVERAL CYTOKINES have been (mRNA) level is increased in postmor- associated with Alzheimer tem AD brains10 and correlates with amy- disease (AD) neuropathol- loid deposition in cerebral blood vessels. ogy. The level of the proin- However, TGF-␤1 may also have noninflammatory cytokine inter- flammatory functions and may play an im- From the Neuroinflammation leukin 6 (IL-6) is increased in the brain, portant role in the growth and survival of S 11-13 Research Laboratories, blood, and cerebrospinal fluid of patients neurons in the AD brain. The expres- 1-6 Department of Psychiatry with AD, and IL-6 has been implicated sion of the cytokine tumor necrosis fac- (Drs Luterman, Yemul, Ho, in the transformation of diffuse to neu- tor ␣ (TNF-␣) is decreased in the frontal and Pasinetti), Department of ritic plaques in the AD brain.7 Interleu- cortex, superior temporal gyrus, and en- Pathology (Dr Purohit), and kin 1 (IL-1) has also been linked to amy- torhinal cortex of AD patients compared Psychiatry (Dr Mohs), Mount loid plaque transition from the diffuse to with non-AD controls14 and has both pro- Sinai School of Medicine, New dense core stage8 and the propagation of tective and destructive functions.15 York, NY; Department of the inflammatory signal through the in- Although cytokine expression has been Psychiatry, Bronx Veterans duction of S100␤.4,9 Transforming growth shown to be regulated in the AD brain, pre- Affairs Medical Center, Bronx, NY (Dr Haroutunain); and factor ␤1 (TGF-␤1) has been shown to vious studies evaluated postmortem tissue Department of Neurology, promote ␤-amyloid deposition in trans- from severely affected patients at the end 10 Georgetown University School genic mouse models and therefore may stages of AD. There is presently little infor- of Medicine, Washington, DC exacerbate amyloidogenic pathology. In mation on the dynamic expression of in- (Dr Aisen). addition, the TGF-␤1 messenger RNA flammatory cytokines relative to the clini-

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1153

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 METHODS complementary DNA (cDNA) templates for the following human cytokine genes was used: IL-6, TGF-␤1, IL-1␤, PATIENT SELECTION CRITERIA TNF-␤, TNF-␣, IL-1␣, IL-1 receptor antagonist (IL-1Ra), and interferon gamma (IFN-␥). The probe set also in- Human postmortem brains from AD and age-matched cluded the housekeeping genes L32 and glyceralde- non-AD cases were obtained from the Alzheimer’s Disease hyde-3Ј phosphate dehydrogenase (GAPDH) for normal- Research Center (ADRC) of the Mount Sinai School of Medi- ization of assay conditions. Details on the generation of cine, New York, NY. The cases selected had either no sig- phosphate P 32–labeled antisense RNA probes and condi- nificant neuropathological features or only neuropatho- tions of the RNase protection assay (RPA) are provided by logical features associated with AD.17,18 A multistep approach the manufacturer. The quantity of radioactively labeled based on cognitive and functional status during the last 6 RNase protection fragments was determined using a Mo- months of life was applied to the assignment of Clinical De- lecular Dynamics Storm 860 Phosphor Screen Scanner with mentia Rating (CDR) scores, as previously reported.17,18 Sub- the ImageQuant software package (Molecular Dynamics, jects were divided into groups on the basis of their CDR Sunnyvale, Calif). Each RPA analysis was conducted with scores as follows: 0, nondemented; 0.5, questionable de- 10 µg of total RNA, according to A260 values. Data are ex- mentia; 1, mild dementia; 2, moderate dementia; and 4 to pressed as a ratio of the specific mRNA of interest normal- 5, very severe dementia. ized to the constitutively expressed GAPDH mRNA. Nor- The extent of neurofibrillary tangle and ␤-amyloid neu- malization of cytokine mRNA signals to L32 mRNA did not ritic plaque deposition was assessed in accord with the Con- change the outcome results (not shown). sortium to Establish a Registry for Alzheimer’s Disease (CERAD) neuropathological battery.17-19 Multiple (5 in gen- RNA SAMPLE QUALITY AND EXCLUSION eral) high-power fields (ϫ200, 0.5 mm2) were examined CRITERIA FOR RPA in each histological slide, containing specimens obtained from multiple brain regions, according to the CERAD re- Initial quality control consisted of agarose gel assessment of gional sampling scheme. The density of neurofibrillary ethidium bromide staining of both 18–Svedberg flotation unit tangles and neuritic plaques was rated on a 4-point scale (Sf) and 28-Sf ribosomal RNA (rRNA) integrity as well as as follows: 0, absent; 1, sparse; 3, moderate; and 5, severe. detection of low-molecular-weight “smearing” of degraded Plaques were visualized following either Bielschowsky sil- mRNA. All samples were subjected to RPA, regardless of ini- ver or thioflavine S staining.20,21 The investigators were blind tial quality control assessment. Only samples with nonde- to the clinical diagnosis of each case until all quantitative graded RNA determined by both ethidium bromide and RPA analyses were completed and values were assigned to each analysis were included in the study. For the entorhinal cor- specimen. tex, 19 (24%) of the 78 subjects were removed from the study because of poor RNA quality; for the superior temporal gy- RNA PREPARATION rus, 16 (20%) of 79 patients were removed; and for the occipital cortex, 9 (36%) of 25 patients were removed. In each Total RNA was prepared with the Ultraspec RNA Isola- brain region examined, RNA exclusion was similarly dis- tion System (Biotecx Laboratories, Houston, Tex), based tributed among the different CDR categories. on the acid guanidinium thiocyanate-phenol-chloroform method.22 STATISTICS

RIBONUCLEASE PROTECTION ASSAY Data were analyzed by analysis of variance (ANOVA) com- bined with a post hoc Dunnett multiple comparison test Total RNA was assayed with the RiboQuant Multiprobe to compare each experimental group with the control group. RNase (ribonuclease) Protection Assay System (PharMin- Correlation analysis was performed by calculating the Pear- gen, San Diego, Calif). A custom probe set containing son R coefficient.

cal progression of AD dementia. The current study is the RESULTS first to examine the mRNA expression patterns of several inflammatory cytokines in different brain regions as a function of the clinical progression of AD. PATIENT POPULATION It has been suggested that the neuropathological features of AD occur in different brain areas in a predict- Patient information for this study is summarized in able temporal pattern during the course of the disease Table 1. Extraneous factors that might influence cyto- progression.16 Therefore, cytokines were measured in the kine gene expression were similar across all groups. Cy- entorhinal cortex (Brodmann area [BM] 36/38) and su- tokine mRNA degradation is a concern with long post- perior temporal gyrus (BM 22), 2 brain areas affected in mortem delays. The ANOVA indicated that there were AD, and in the occipital cortex (BM 17), an area that is no significant differences (PϾ.05) among the different less affected in AD. In this study, we determined the pat- CDR groups with respect to postmortem intervals (data tern of cytokine gene expression as a function of AD clini- not shown). Furthermore, all groups had a statistically cal dementia and the neuropathological features of AD similar (PϾ.05) age at death (data not shown). Cyto- and discussed the relevance of these findings to the de- kine expression can also be impacted by antemortem sign of anti-inflammatory drug trials in AD. inflammatory conditions that may require anti-

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1154

Mean Median CERAD Rating CDR No. of Postmortem Mean No. (%) Score Patients Interval, min Age, y Female Plaques Tangles Entorhinal Cortex (BM 36) 0 13 537.9 83.0 11 (85) 0 1 0.5 8 522.1 88.5 8 (100) 1 5 1 18 259.1 86.7 13 (72) 1 3 2 14 287.9 89.1 10 (71) 3 5 4 2 400.0 78.0 1 (50) 4 5 5 6 191.7 82.5 4 (67) 3 5 Superior Temporal Gyrus (BM 22) 0 14 521.6 83.6 12 (86) 0 0 0.5 10 454.7 88.4 8 (80) 1 0 1 19 282.8 88.4 4 (74) 3 0 2 12 321.8 88.8 1 (92) 4 2 4 2 400.0 78.0 1 (50) 5 4 5 7 433.6 83.9 6 (86) 5 5 Occipital Cortex (BM 17) 0 8 639.4 85.5 8 (100) ...... 4 2 322.5 87.5 1 (50) ...... 5 6 198.0 86.3 5 (83) ......

*CDR indicates Clinical Dementia Rating; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; BM, Brodmann area; and ellipses, not applicable.

inflammatory drug administration. There were a few pa- Clinical Dementia Rating Score tients in the study who had evidence of antemortem non- 00.511245 steroid anti-inflammatory drug/glucocorticoid use, and Undigested Protected Probes Fragments their cytokine profiles did not statistically differ (PϾ.05) β from those of noninflammatory cohorts (data not shown). TNF- Cause of death was considered for all the patients to rule out the possibility that infection or other inflammatory TNF-α events might impact the cytokine levels measured in the IL-1α study. Of the 64 patients included in the study, 51 died IL-1β from acute “cardiac failure” (including cardiorespira- IL-1β IL-1Ra tory failure/arrest, ventricular fibrillation, cardiopulmo- nary arrest, myocardial infarct, and congestive heart fail- IL-6 ure), 2 from cancer, 2 from heart disease (specified), 1 from pneumonia, and 8 from unknown cause. There- TGF-β1 IL-6 fore, long-term inflammatory events are not a confound- IFN-γ ing factor in the study. TGF-β1

BASAL CYTOKINE LEVELS IN COGNITIVELY L32 NORMAL PATIENTS

Total RNA was isolated from the entorhinal cortex, su- GAPDH perior temporal gyrus, and occipital cortex and subjected to RPA analysis. Results of a representative RPA L32 analysis are presented in Figure 1. For TNF-␤, TNF-␣, ␣ ␥ IL-1 , IL-1Ra, and IFN- , mRNA expression was below GAPDH the level of detection for this assay system (Table 2 and Figure 1). Levels of IL-6, TGF-␤1, and IL-1␤ mRNA were Figure 1. Representative ribonuclease protection assay gel. In each lane, determined for the 3 brain regions and expressed as a ra- total RNA (10 µg) from the entorhinal cortex of a cognitively normal control tio to mRNA from the housekeeping gene GAPDH (Table case (Clinical Demential Rating [CDR] score, 0) and patients with 2, Figure 1, and Figure 2). questionable (CDR, 0.5), mild (CDR, 1), moderate (CDR, 2), severe (CDR, 4), and very severe (CDR, 5) clinical dementia were hybridized with Expression of IL-6 mRNA was greater than 10-fold complementary RNA probes. TNF indicates tumor necrosis factor; IL, higher in the entorhinal cortex and superior temporal gy- interleukin; TGF, transforming growth factor; IFN, interferon; and GAPDH, rus as compared with the occipital cortex (Figure 2, A). glyceraldehyde-3Ј phosphate dehydrogenase. In contrast, expression of TGF-␤1 mRNA was relatively lower than that of IL-6 mRNA and was consistent across mRNA was of low abundance and showed no regional all 3 brain regions (Figure 2, B). Expression of IL-1␤ differences (Figure 2, C).

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1155

Entorhinal Cortex Superior Temporal Occipital Cortex n = 14 (BM 36/38) Gyrus (BM 22) (BM 17) Cytokine (n = 13) (n = 14) (n=8) 0.050 n = 13

IL-6 0.041 ± 0.005 0.050 ± 0.006 0.004 ± 0.001 mRNA TGF-␤1 0.023 ± 0.003 0.015 ± 0.002 0.018 ± 0.004

IL-1␤ 0.005 ± 0.001 0.004 ± 0.001 0.006 ± 0.002 GAPDH TNF-␤ ND ND ND 0.025 TNF-␣ ND ND ND IL-6/ IL-1␣ ND ND ND

IL-1Ra ND ND ND n = 8 IFN-␥ ND ND ND 0.000 Entorhinal Cortex Superior Temporal Gyrus Occipital Cortex *Values are ratios and are expressed as mean ±SEM. CDR indicates Clinical Dementia Rating; mRNA, messenger RNA; GAPDH, B glyceraldehyde-3Ј phosphodehydrogenase; BM, Brodmann area; IL, 0.075 TGF-β1 interleukin; TGF, transforming growth factor; TNF, tumor necrosis factor; ND, not detectable under standard assay conditions; and IFN, interferon.

INCREASED CYTOKINE mRNA EXPRESSION 0.050 IN ENTORHINAL CORTEX AND SUPERIOR mRNA TEMPORAL GYRUS AS A FUNCTION GAPDH

1/ n = 13

OF CDR SCORE β 0.025 n = 8

TGF- n = 14 IL-6 mRNA Expression

Interleukin 6 mRNA expression was measured in the en- 0.000 torhinal cortex, superior temporal gyrus, and occipital cor- Entorhinal Cortex Superior Temporal Gyrus Occipital Cortex tex from cases at different stages of AD dementia as as- C sessed by the CDR score (Figure 3, A-C). For the 0.075 IL-1β entorhinal cortex, there was an overall difference in IL-6 mRNA expression across all CDR groups examined (ANOVA: PϽ.001, F5,60=8.3). In addition, cases characterized by severe/terminal dementia (CDR, 5) had signifi- 0.050

cantly elevated IL-6 mRNA levels (PϽ.01; Figure 3, A) as mRNA compared with cognitively normal patients (CDR, 0).

GAPDH

In the superior temporal gyrus, as in the entorhinal / β cortex, there was also an overall difference in IL-6 mRNA 0.025 expression across the CDR groups (ANOVA: P=.03, IL-1 F5,63=2.7). Cases characterized by severe/terminal de- n = 8 mentia (CDR, 5) had significantly elevated IL-6 mRNA n = 13 n = 14 levels (PϽ.01; Figure 3, B) as compared with cogni- 0.000 tively normal cases (CDR, 0). Entorhinal Cortex Superior Temporal Gyrus Occipital Cortex No detectable change in IL-6 mRNA expression was Figure 2. Regional distribution of basal cytokine messenger RNA (mRNA) found in the occipital cortex (Figure 3, C). expression in cases characterized by normal cognitive status (Clinical Dementia Rating score, 0). Data are expresed as a ratio of expression of the mRNA of interest normalized to GAPDH mRNA. Values are mean ± SEM. See TGF-␤1 mRNA Expression Figure 1 legend for expansions of additional abbreviations.

A pattern similar to that of IL-6 mRNA expression was found for TGF-␤1 mRNA expression in the 3 brain re- significantly elevated TGF-␤1 mRNA levels (PϽ.01; Fig- gions examined (Figure 3, D-F). Although TGF-␤1 mRNA ure 3, E) as compared with cognitively normal cases levels in the entorhinal cortex of cases characterized by (CDR, 0). severe/terminal dementia (CDR, 5) increased greater than There was no change in TGF-␤1 mRNA expression 1-fold relative to cognitively normal cases (CDR, 0), the in the occipital gyrus of patients who were very severely change was not statistically significant by ANOVA demented (CDR, 5; Figure 3, F). (F5,60=1.948; P=.01) (Figure 3, D). In both the entorhinal cortex and superior tempo- Similar to the results for IL-6 mRNA expression, in ral gyrus, TGF-␤1 and IL-6 mRNA expression in cases the superior temporal gyrus, there was an overall differ- characterized by questionable (CDR, 0.5), mild (CDR, ence in TGF-␤1 mRNA expression among the different 1), moderate (CDR, 2), or severe (CDR, 4) dementia did CDR groups (ANOVA: P=.009, F5,63=3.4), and cases not differ from cognitively normal (CDR, 0) cases (Fig- characterized by severe/terminal dementia (CDR, 5) had ure 3, E and F).

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1156

0.10 Entorhinal Cortex 0.075 Entorhinal Cortex *P<.01 vs CDR Score of 0

* 0.050 mRNA mRNA 0.05

GAPDH

GAPDH 1/ β

IL-6/ 0.025 TGF-

0.00 0.000 010.5 2 4 5 010.5 2 4 5 CDR Score CDR Score B E 0.3 Superior Temporal Gyrus 0.100 Superior Temporal Gyrus *P<.01 vs CDR Score of 0 *P<.01 vs CDR Score of 0

0.075 0.2 mRNA mRNA 0.050

GAPDH

GAPDH 1/ β *

IL-6/ 0.1

* TGF- 0.025

0.0 0.000 010.5 2 4 5 010.5 2 4 5 CDR Score CDR Score C F 0.10 Occipital Cortex 0.05 Occipital Cortex

0.04

mRNA 0.03 mRNA 0.05

GAPDH

GAPDH 1/

β 0.02 IL-6/ TGF-

0.01

0.00 0.00 05 05 CDR Score CDR Score Figure 3. Increasing IL-6 and TGF-␤1 messenger RNA (mRNA) expression in the Alzheimer disease brain as a function of CDR score. Data are expressed as a ratio of expression of the mRNA of interest normalized to GAPDH mRNA. Horizontal line indicates mean. See Figure 1 legend for expansions of additional abbreviations.

IL-1␤ mRNA Expression CYTOKINE EXPRESSION AS A FUNCTION OF AD NEUROPATHOLOGY (CERAD) Interleukin 1␤ mRNA expression was very low in all brain areas examined and did not change as a function of CDR IL-6 mRNA Expression score (data not shown). In the entorhinal cortex, there was an overall difference Correlation of IL-6 and TGF-␤1 mRNA Levels in IL-6 mRNA expression across groups stratified by CERAD neurofibrillary tangle rating (ANOVA: P=.04, In the entorhinal cortex, IL-6 and TGF-␤1 mRNA levels F2,60=3.4). In addition, patients with severe neurofibril- showed a high degree of correlation (Pearson R, 0.64; lary tangles (CERAD, 5) had significantly elevated IL-6 PϽ.001). This correlation was more pronounced in the mRNA expression (PϽ.05; Figure 4, A) as compared superior temporal gyrus (Pearson R, 0.84; PϽ.001). with patients with absent or sparse neurofibrillary tangles

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1157

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 Entorhinal Cortex n = 30 n = 16 A 0.05 *P<.05 vs CERAD Score of 0-1 * B 0.075 n = 8 0.04 n = 15 0.050 n = 17 n = 18 0.03 n = 18 mRNA mRNA

GAPDH GAPDH 0.02 0.025 IL-6/ IL-6/ 0.01

0.00 0.000 0-1 3 5 0135 Neurofibrillary Tangles (CERAD) Neuritic Plaques (CERAD) Superior Temporal Gyrus

C 0.15 *P<.01 vs CERAD Score of 0-1 D 0.100 n = 7 n = 19 *

0.075 0.10 n = 18 n = 18 n = 9 mRNA mRNA n = 8 0.050 n = 49 GAPDH GAPDH 0.05 IL-6/ IL-6/ 0.025

0.00 0.000 0-1 3 5 0135 Neurofibrillary Tangles (CERAD) Neuritic Plaques (CERAD) Entorhinal Cortex

E 0.04 F 0.04

n = 8 n = 30 0.03 0.03 n = 17 n = 18 n = 16 n = 18 n = 15 mRNA mRNA

0.02 0.02

GAPDH GAPDH 1/ 1/ β β TGF-

TGF- 0.01 0.01

0.00 0.00 0-1 3 5 0135 Neurofibrillary Tangles (CERAD) Neuritic Plaques (CERAD) Superior Temporal Gyrus

G 0.02 *P<.01 vs CERAD Score of 0-1 H 0.04 n = 49 n = 19 0.03 mRNA mRNA n = 18 0.01 0.02 n = 9

n = 8* n = 7 GAPDH n = 18 GAPDH * 1/ 1/ β β TGF-

TGF- 0.01

0.00 0.00 0-1 3 5 0135 Neurofibrillary Tangles (CERAD) Neuritic Plaques (CERAD)

Figure 4. Changes in IL-6 and TGF-␤1 messenger RNA (mRNA) expression in the Alzheimer disease brain as a function of the severity of neurofibrillary tangle and neurotic plaque pathological features rated according to a 4-point scale based on the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) regional sampling scheme. Data are expressed as a ratio of expression of the mRNA of interest normalized to GAPDH mRNA. Values are mean ± SEM. See Figure 1 legend for expansions of additional abbreviations.

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1158

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 (CERAD, 0-1). There was no change in IL-6 mRNA ex- with increased levels of neurofibrillary tangles in the en- pression with increasing neuritic plaques (Figure 4, B). torhinal cortex and superior temporal gyrus, the TGF-␤1 This same pattern of changes was found in the mRNA level was inversely associated with neurofibril- superior temporal gyrus, with an overall difference in lary tangles in the superior temporal gyrus. IL-6 mRNA expression across groups stratified by This study suggests that cytokine expression in the CERAD neurofibrillary tangle rating (ANOVA: P=.002, brain may play an important role as a conditional factor F2,63=6.7). As observed in the entorhinal cortex, cases for neurodegenerative events in the later stages of clini- with severe neurofibrillary tangles (CERAD, 5) had sig- cal disease. Our data are consistent with a localized in- nificantly elevated IL-6 mRNA expression (PϽ.01; Fig- flammatory response to late AD neurodegeneration but ure 4, C) compared with cases with absent or sparse do not support a role for cytokine-mediated destructive neurofibrillary tangles (CERAD, 0-1). No detectable inflammation in the early progression of AD. change in IL-6 mRNA expression with increasing neu- Interleukin 6 mRNA expression was measured from ritic plaque levels was found in the superior temporal the same group of patients after stratification by either gyrus (Figure 4, D). CDR or CERAD neuropathological ratings. Elevation of IL-6 mRNA was observed in cases defined by severe/ TGF-␤1 mRNA Expression terminal dementia or high levels of neurofibrillary tangles in the entorhinal cortex or superior temporal gyrus. The Transforming growth factor ␤1 mRNA expression in the fact that IL-6 mRNA expression was increased in severe entorhinal cortex did not change with increasing neu- AD cases, defined by either dementia or neuropathologi- rofibrillary tangles or neuritic plaques (Figure 4, E and cal features, further supports the use of the CDR scale F, respectively). to assess disease progression and confirms previous stud- In the superior temporal gyrus, there was an over- ies correlating neuropathological features with CDR.17,18 all inverse relationship (ANOVA: PϽ.001, F2,63=12.2) be- Like IL-6, TGF-␤1 mRNA levels were increased in tween TGF-␤1 mRNA expression and neurofibrillary the superior temporal gyrus and entorhinal cortex of se- tangles. Cases with moderate to severe neurofibrillary verely demented cases. In contrast to IL-6, however, tangle pathological features (CERAD, 3 or 5) had sig- TGF-␤1 mRNA expression decreased in the superior tem- nificantly lower TGF-␤1 mRNA expression (PϽ.01; Fig- poral gyrus as a function of neurofibrillary tangle levels. ure 4, G) as compared with patients with absent or sparse This suggests that the 2 cytokines may be linked to dif- neurofibrillary tangles (CERAD, 0-1). There was no re- ferent neuropathological mechanisms and that there may lationship between TGF-␤1 mRNA expression and neu- be qualitatively different mechanisms active in the 2 brain ritic plaques (Figure 4, H). regions. Previous studies23 have shown that TGF-␤1 can be produced by both neurons and glia. Ongoing immu- Because there are relatively few neuropathological nocytochemical and in situ hybridization studies in our changes caused by AD in the occipital cortex, the corre- laboratory will further clarify the cell-type expression of sponding neurofibrillary tangle and neuritic plaque data TGF-␤1 and other cytokines during the clinical progres- are unavailable. sion of AD dementia. Expression of TNF-␤, TNF-␣, IL-1␣, IL-1Ra, and Previous studies8,9 have suggested that microglial IL-1 IFN-␥ mRNAs were all below the level of detection for expression is up-regulated in the AD brain and support this assay system (Table 2 and Figure 1) and did not the hypothesis that IL-1 may be involved in the matura- change with dementia or neuropathological features (data tion of neuritic plaques. In the present study, we found not shown). no elevation of IL-1␤ mRNA expression during the clinical progression of AD. This apparent conflict may be re- COMMENT lated to methodological issues. Our study is a broad screen- ing of cases at various levels of cognitive dysfunction using Inflammatory cytokine gene expression was measured in the RPA technique to measure cytokine mRNA expres- 3 different brain regions from cases characterized by dif- sion. The earlier studies used primarily immunocyto- ferent stages of AD clinical dementia (CDR) and differ- chemical techniques, which cannot easily be scaled up to ent levels of neuritic plaques and neurofibrillary tangles. the level used in our study and may not be reliable indi- For TNF-␤, TNF-␣, IL-1␣, IL-1Ra, and IFN-␥, mRNA ex- cators of cytokine expression (particularly in view of non- pression was below the level of detection for our assay specific adsorption of many molecules to plaques). Fur- system in the cognitively normal brain and did not change thermore, careful clinical staging of cases was not available with dementia or increasing plaques and tangles. In cases in the earlier studies. Moreover, recent data show that characterized by normal cognitive status, we found that TGF-␤1 knockout mice have elevated IL-1␤ mRNA lev- IL-6 mRNA expression was greater than 10-fold higher els,24 suggesting that the 2 cytokines may have a recipro- in the entorhinal cortex and superior temporal gyrus rela- cal relationship. The lack of IL-1␤ mRNA in our study may tive to the occipital cortex. The level of IL-6 and TGF-␤1 be related to the increases we found in TGF-␤1 mRNA. mRNA was increased only in the entorhinal cortex and The present data suggest that inflammatory cytokine superior temporal gyrus of cases characterized by severe/ genes are not up-regulated in early AD. Moreover, the lack terminal clinical dementia (CDR, 5) but not question- of a correlation between IL-6 mRNA expression and neu- able, mild, or moderate dementia. Furthermore, IL-6 and ritic plaque density challenges the hypothesis that amy- TGF-␤1 mRNA levels were highly correlated with each loid deposition initiates the cerebral acute-phase response other. In contrast, while the IL-6 mRNA level correlated in AD.2 Surprisingly, this study suggests that IL-6 mRNA

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1159

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 expression may be related to neurofibrillary tangle levels 2. Gruol DL, Nelson TE. Physiological and pathological roles of interleukin-6 in the by a mechanism that is not presently understood. central nervous system. Mol Neurobiol. 1997;15:307-339. 3. Kalman J, Juhasz A, Laird G, et al. Serum interleukin-6 levels correlate with the It has been suggested that drugs that suppress cy- severity of dementia in Down syndrome and in Alzheimer’s disease. Acta Neurol tokine expression, such as glucocorticoids, may be ben- Scand. 1997;96:236-240. eficial in the treatment of AD. However, the absence of 4. Singh VK. Studies of neuroimmune markers in Alzheimer’s disease. Mol Neu- inflammatory cytokine up-regulation in early to moder- robiol. 1994;9:73-81. 5. Singh VK, Guthikonda P. Circulating cytokines in Alzheimer’s disease. J Psychi- ate stages of AD dementia found in this study does not atr Res. 1997;31:657-660. support a central role for such mediators in disease pro- 6. Terreni L, De Simoni MG. Role of the brain in interleukin-6 modulation. Neuro- gression. Indeed, this hypothesis is consistent with the immunomodulation. 1998;5:214-219. failure of prednisone therapy in the treatment of mild to 7. Hull M, Berger M, Volk B, Bauer J. Occurrence of interleukin-6 in cortical plaques moderate AD.25 It may be more appropriate to direct anti- of Alzheimer’s disease patients may precede transformation of diffuse into neuritic plaques. Ann N Y Acad Sci. 1996;777:205-212. inflammatory treatment strategies at target molecules that 8. Sheng JG, Mrak RE, Griffin WS. Neuritic plaque evolution in Alzheimer’s disease seem to be involved in the early stages of the disease, such is accompanied by transition of activated microglia from primed to enlarged to as those found for neuronal cyclooxygenase 2 (Ho et al, phagocytic forms. Acta Neuropathol (Berl). 1997;94:1-5. personal communication). 9. Griffin WS, Sheng JG, Royston MC, et al. Glial-neuronal interactions in Alzhei- mer’s disease: the potential role of a ‘cytokine cycle’ in disease progression. Brain Much effort has been directed toward clarifying the Pathol. 1998;8:65-72. role of inflammatory cytokines in neurodegenerative dis- 10. Wyss-Coray T, Masliah E, Mallory M, et al. Amyloidogenic role of cytokine TGF- orders such as AD. Careful characterization of the tem- beta1 in transgenic mice and in Alzheimer’s disease. Nature. 1997;389:603-606. poral and regional regulation of cytokine expression is 11. Lippa CF, Smith TW, Flanders KC. Transforming growth factor-beta: neuronal and necessary to further elucidate cytokine mechanisms and glial expression in CNS degenerative diseases. Neurodegeneration. 1995;4:425-432. 12. Pratt BM, McPherson JM. TGF-beta in the central nervous system: potential roles the potential of anticytokine therapy. In addition, be- in ischemic injury and neurodegenerative diseases. Cytokine Growth Factor Rev. cause mRNA levels do not necessarily reflect changes in 1997;8:267-292. bioactivity, future studies will attempt to explore these 13. Flanders K, Ren RF, Lippa CF. Transforming growth factor-betas in neurodegen- relationships at the protein level. Nonetheless, the erative disease. Prog Neurobiol. 1998;54:71-85. 14. Lanzrein AS, Johnston CM, Perry VH, Jobst KA, King EM, Smith AD. Longitudinal present studies reveal an interesting regional pattern of study of inflammatory factors in serum, cerebrospinal fluid, and brain tissue in Alz- cytokine gene expression in the AD brain that may be heimer’s disease: interleukin-1 beta, interleukin-6, interleukin-1 receptor antagonist, relevant to neurodegenerative disease but do not sup- tumor necrosis factor-alpha, the soluble tumor necrosis factor receptors I and II, port the theory that cytokines are intimately involved in and alpha1-antichymotrypsin. Alzheimer Dis Assoc Disord. 1998;12:215-227. plaque maturation in early AD or other processes cen- 15. Mattson MP, Barger SW, Furukawa K, et al. Cellular signaling roles of TGF beta, TNF alpha and beta APP in brain injury responses and Alzheimer’s disease. Brain tral to the progression of early AD. Res Brain Res Rev. 1997;23:47-61. 16. Braak H, Braak E. Evolution of neuronal changes in the course of Alzheimer’s Accepted for publication March 13, 2000. disease. J Neural Transm Suppl. 1998;53:127-140. This work was supported by grants AG13799, AG14329, 17. Haroutunian V, Purohit DP, Perl DP, et al. Neurofibrillary tangles in nondemented and AG16743 (Dr Pasinetti), AG05138 (Alzheimer Disease elderly subjects and mild Alzheimer disease. Arch Neurol. 1999;56:713-718. 18. Haroutunian V, Perl DP, Purohit DP, et al. Regional distribution of neuritic plaques Research Center [ADRC] of Mount Sinai) (Dr Davis), and in nondemented elderly and cases of very mild Alzheimer disease. Arch Neurol. AG02219 (program project) (Dr Mohs), National Institute 1998;55:1185-1191. on Aging, National Institutes of Health, Bethesda, Md. 19. Mirra SS. The CERAD neuropathology protocol and consensus recommenda- The authors thank Daniel Perl, MD, Mount Sinai Alz- tions for the postmortem diagnosis of Alzheimer’s disease: a commentary. Neu- robiol Aging. 1997;18(4 suppl):S91-S94. heimer’s Disease Research Center, New York, NY, for neu- 20. Bancroft JD, Stevens A. Theory and Practice of Histological Techniques. New ropathological support. York, NY: Churchill Livingstone Inc, 1997. Corresponding author: Giulio Maria Pasinetti, MD, 21. Yamamoto T, Hirano A. A comparative study of modified Bielschowsky, Bodian PhD, Neuroinflammation Research Laboratories, Depart- and thioflavin S stains on Alzheimer’s neurofibrillary tangles. Neuropathol Appl ment of Psychiatry, Box 1229, Mount Sinai Medical Neurobiol. 1986;12:3-9. 22. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium Center, One Gustave L Levy Place, New York, NY 10029 thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156-159. (e-mail: [email protected]). 23. Finch CE, Laping NJ, Morgan TE, Nichols NR, Pasinetti GM. TGF-beta1 is an or- ganizer of responses to neurodegeneration. J Cell Biochem. 1993;53:314-322. 24. Saito H, Shultz LD, Sinha M, Papaconstantinou J. Induction of the alpha(1)- REFERENCES antichymotrypsin gene in the brain associated with TGF-beta1 deficiency or sys- temic administration of endotoxin. Biochem Biophys Res Commun. 1999;263: 1. Bonaccorso S, Lin A, Song C, et al. Serotonin-immune interactions in elderly vol- 270-275. unteers and in patients with Alzheimer’s disease (DAT): lower plasma trypto- 25. Aisen PS, Davis KL, Berg J, et al, for the Members of the Alzheimer’s Disease phan availability to the brain in the elderly and increased serum interleukin-6 in Cooperative Study. A randomized controlled trial of prednisone in Alzheimer’s DAT. Aging (Milano). 1998;10:316-323. disease. Neurology. 2000;54:588-593.

(REPRINTED) ARCH NEUROL / VOL 57, AUG 2000 WWW.ARCHNEUROL.COM 1160