Functional Imbalance of Anterolateral Entorhinal Cortex and Hippocampal Dentate/CA3 Underlies Age-Related Object Pattern Separation Deficits

Abstract

The entorhinal cortex (EC) is among the earliest brain areas to deteriorate in Alzheimer's disease (AD). However, the extent to which functional properties of the EC are altered in the aging brain, even in the absence of clinical symptoms, is not understood. Recent human fMRI studies have identified a functional dissociation within the EC, similar to what is found in rodents. Here, we used high-resolution fMRI to identify a specific hypoactivity in the anterolateral EC (alEC) commensurate with major behavioral deficits on an object pattern separation task in asymptomatic older adults. Only subtle deficits were found in a comparable spatial condition, with no associated differences in posteromedial EC between young and older adults. We additionally linked this condition to dentate/CA3 hyperactivity, and the ratio of activity between the regions was associated with object mnemonic discrimination impairment. These results provide novel evidence of alEC-dentate/CA3 circuit dysfunction in cognitively normal aged humans.

Keywords: CA3; LEC; aging; alEC; dentate gyrus; entorhinal; fMRI; lateral entohrinal; memory; pattern separation.

Figure 1. Task schematic and behavioral performance

A) Illustrative diagram of an object block (3 blocks completed). B) Illustrative diagram of a spatial block (3 blocks completed). Objects were smaller relative to screen size in the actual task, and presentation order was randomized across runs (i.e., study and test orders were different) and across participants. Stimuli were presented for 3 seconds, with a 1 second inter-stimulus interval. C) Target hit rates across test domains. No differences were observed. D) Lure discrimination rates across test domains. Older subjects were significantly impaired at object, but not spatial discrimination overall. E) Lure discrimination across the five similarity bins. Whereas older subjects were relatively impaired at discrimination of mid-similarity spatial lures, they were more globally impaired at object discrimination. Data are shown as mean ± standard error, with each point representing a single subject in panels c and d. (* = Young > Old in panel d; + = Young > Old for object lures in panel E; # = Young > Old for spatial lures in panel E; all post hoc tests are reported as p < 0.05 corrected for multiple comparisons.)

Figure 2. Regions of interest (ROIs)

Hippocampal ROIs were based on prior studies^12,13,24,40, and segmentation of alEC and pmEC was done by hand in accordance with the atlas generated by Maass and colleagues⁸. (PRC = perirhinal cortex; aLEC = anterolateral entorhinal cortex; pMEC = posteromedial entorhinal cortex; PHC = parahippocampal cortex; Sub = Subiculum; DG = dentate gyrus.)

Figure 3. Task and age effects in left alEC and right pmEC

A) Collapsed across similarity levels, older participants show significantly lower activity in left alEC during object discrimination. B) Older adults show significantly lesser gains in alEC engagement with decreasing object similarity compared to young participants. C) alEC shows neither a modulation of spatial lure similarity nor age during spatial discrimination. D) Collapsed across similarity levels, both old and young participants show comparable levels of right pmEC engagement during spatial lure discrimination. E) pmEC shows neither a modulation of object lure similarity nor age during object discrimination. F) For both young and old participants, pmEC is increasingly engaged as lure similarity decreases during spatial discrimination, with no group difference. Data are shown as mean ± standard error (CR = correct rejection; * = Young > Old; + = Significantly different group slopes for the curve of alEC engagement across similarity levels; all post hoc tests are reported as p < 0.05 corrected for multiple comparisons.)

Figure 4. Task and age effects in left DG/CA3 and CA1

A) Collapsed across similarity levels, older participants show significantly higher activity in left DG/CA3 across both test domains. B, C) Compared to young participants, older adults show significantly greater activity in left DG/CA3 across all lure similarity levels during object and spatial discrimination. D, E) Both young and older participants show significantly increasing engagement of left CA1 with decreasing lure similarity, but no group differences. Data are shown as mean ± standard error. (CR = correct rejection; * = Young > Old; # = Significantly different group intercepts (but not slopes) for curves of DG/CA3 engagement across similarity levels; all post hoc tests are reported as p < 0.05 corrected for multiple comparisons.)

Figure 5. Correlations between left DG/CA3, left alEC, and behavior

A, B) Left DG/CA3 engagement is negatively correlated with object and spatial discrimination performance across participants. The correlation was only significant for older participants during object discrimination, whereas for spatial discrimination, the correlation was significant across the entire sample but not for either age group individually. C) Left alEC engagement is positively correlated with object discrimination performance. Though the relationship is qualitatively similar for young participants, it reaches significance only for older adults. (Solid regression lines indicate significant whole-sample correlations, whereas broken regression lines indicate significance for a given age group.)

Figure 6. Normed ratios of left DG/CA3 vs. alEC activity across groups and relationship with behavior

A) Older participants have a significantly higher ratio value than young adults, indicating greater DG/CA3 activity relative to alEC. Conversely, young participants have a near-zero mean ratio, indicating relatively balanced engagement of both regions. B) Object discrimination performance negatively correlates with this normed ratio in older adults, indicating greater DG/CA3 activity relative to alEC is associated with poorer object mnemonic discrimination. Despite a qualitatively similar relationship in young participants, the relationship does not yield a significant correlation. (Solid regression lines indicate significant whole-sample correlations, whereas broken regression lines indicate significance for a given age group).

Figure 7. Cross-regional correlations across conditions and age groups

During object trials: A) Young participants feature significant correlations between alEC, PRC, DG/CA3, and CA1. B) Older participants feature significant correlations between alEC and PRC, alEC and PHC, and between PHC and CA1. C) Young participants have significantly higher correlations between DG/CA3 and alEC, DG/CA3 and PRC, and between DG/CA3 and CA1 compared to older participants. During spatial trials: D) Young participants feature significant correlations between pmEC, PHC, DG/CA3, and CA1, as well as between alEC and pmEC. E) Older participants feature significant correlations between pmEC and PHC, and between alEC and pmEC. F) Young participants have significantly higher correlations between DG/CA3 and CA1 compared to older participants (there is a trend toward age-related decreases in CA1 and both pmEC and PHC, but this difference is not significant). (* = significant correlation for left and middle panels, and significant group difference for right panels.)