Our previous studies have implicated perlecan, a specific heparan sulfate proteoglycan, in the pathogenesis of fibrillar beta-amyloid protein (A beta) accumulation and persistence in Alzheimer's disease (AD) brain. In the present investigation, we determined if perlecan mRNA was present in rodent and human brain tissue and whether perlecan persistence in A beta amyloid deposits in AD hippocampus may be partly due to increased perlecan expression and/or decreased perlecan degradation. To detect and to quantify low-abundance perlecan mRNA in rodent and postmortem human brain tissue, regions of perlecan domain I (503 and 366 bp) containing the unique heparan sulfate glycosaminoglycan attachment sites were analyzed by reverse transcription (RT) and polymerase chain reaction (PCR). Perlecan mRNA was detected in rodent brain, kidney, and liver and in human AD and normal aged frontal cortex. Different-size transcripts of perlecan domain I were found, suggesting the existence of alternatively spliced variants of perlecan or closely related gene products. Quantitative competitive RT-PCR using a mutant perlecan domain I internal standard was used to determine perlecan mRNA levels in total RNA isolated from the hippocampus of 10 AD (mean +/- SEM duration of illness, 11.3 +/- 1.4 years) and 10 normal aged controls. No significant difference in perlecan mRNA levels from the hippocampus of AD (1.12 +/- 0.29 amol/500 ng of total RNA) versus normal aged controls (1.09 +/- 0.30 amol/500 ng of total RNA) was found, indicating that perlecan expression remained at steady-state levels. These results therefore suggest that perlecan persistence in A beta-amyloid deposits in late-stage AD may be primarily due to decreased perlecan degradation and removal.