Analysis of Mitochondrial haemoglobin in Parkinson's disease brain

Abstract

Mitochondrial dysfunction is an early feature of neurodegeneration. We have shown there are mitochondrial haemoglobin changes with age and neurodegeneration. We hypothesised that altered physiological processes are associated with recruitment and localisation of haemoglobin to these organelles. To confirm a dynamic localisation of haemoglobin we exposed Drosophila melanogaster to cyclical hypoxia with recovery. With a single cycle of hypoxia and recovery we found a relative accumulation of haemoglobin in the mitochondria compared with the cytosol. An additional cycle of hypoxia and recovery led to a significant increase of mitochondrial haemoglobin (p<0.05). We quantified ratios of human mitochondrial haemoglobin in 30 Parkinson's and matched control human post-mortem brains. Relative mitochondrial/cytosolic quantities of haemoglobin were obtained for the cortical region, substantia nigra and cerebellum. In age matched post-mortem brain mitochondrial haemoglobin ratios change, decreasing with disease duration in female cerebellum samples (n=7). The change is less discernible in male cerebellum (n=18). In cerebellar mitochondria, haemoglobin localisation in males with long disease duration shifts from the intermembrane space to the outer membrane of the organelle. These new data illustrate dynamic localisation of mitochondrial haemoglobin within the cell. Mitochondrial haemoglobin should be considered in the context of gender differences characterised in Parkinson's disease. It has been postulated that cerebellar circuitry may be activated to play a protective role in individuals with Parkinson's. The changing localisation of intracellular haemoglobin in response to hypoxia presents a novel pathway to delineate the role of the cerebellum in Parkinson's disease.

Keywords: Cerebellum; Gender; Haemoglobin; Hypoxia; Mitochondria; Parkinson's disease.

Supplementary Fig. 1b

All western blots used to calculate fly Hb under cyclical hypoxia.

Supplementary Fig. 3

Ratio of mitochondrial versus cytoplasmic HbB in controls and Parkinson's cerebellum with increase in PMI. Control samples show a general decrease in HbB. The lower plot shows values for male and female HbB and is also coded for PMI category illustrating the lack of relationship between PMI and HbB ratios.

Supplementary Fig. 4

Levels of COXIV a mitochondrial marker, plotted against PMI. Supplementary Fig. 6. Equal numbers (n = 7 + 7) of male and female samples age-matched with similar disease duration also show a reduction in cerebellar HbB in the mitochondrial fraction from female cerebellum.

Supplementary Fig. 5

Alpha synuclein categories plotted for each group of Parkinson's brains.

Supplementary Fig. 6

Haemoglobin ratios plotted for equal numbers of male and female age/disease matched brains.

Supplemental Fig. 7

Fig. 1

A. Mitochondrial HbA migrates from the intermembrane space to the outer membrane in affected human male cerebellum. Mitochondrial samples were sub-fractioned to allow examination of HbA localisation within the organelle, a representative gel is shown for each gender and control (total n = 8). The male Parkinson's disease brain demonstrated a shift in HbA localisation from the intermembrane space to the outer membrane fraction. This was not seen in the control or female Parkinson's brain mitochondria. Levels of HbA in the IMS were quantified in control and PD samples for both male and female patients (n = 2 for each), using Image J. Please see Supplemental Fig. 7 for all gel images. Levels of HbA in the IMS were significantly decreased in male PD compared with male control (p = 0.028 using unpaired two-tailed t-test). No significant change in female PD compared with female control (p $_amp_$gt; 0.05, unpaired two-tailed t-test). Mito fr – mitochondrial fraction, OM – outer membrane, IMS – inter membrane space, IM – inner membrane, M - matrix. B. Cycles of hypoxia result in increased hb in Drosophila mitochondrial fractions. Mitochondrial/cytoplasmic Hb levels determined using Western blotting, normalised to beta-actin. Hypoxia conditions: 2.5% O₂ 30 min 25 °C followed by normoxia 30 min 25 °C (middle bar) 2.5% O₂ 30 min 25 °C followed by normoxia 30 min 25 °C × 2 (right hand bar). 40–100 flies per condition. n = 3, * p $_amp_$lt; 0.05 (1 tailed t-test).

Fig. 2

Representative gel images for mitochondrial/cytoplasmic Hb ratios - normalised to beta actin, were determined by Western blotting. COXIV antibody was utilised for quality control of mitochondrial versus cytoplasmic fractions. To summarise data, samples were grouped into age ranges (1 $_amp_$lt; 70, 2 70–79, 3 ≥ 80) and separated by sex and diagnosis type (control, late PD: late stage disease diagnosed over the age of 60, early PD: early stage disease diagnosed over the age of 60, young PD: early onset disease diagnosed under the age of 60). Values were visualised as boxplots. Boxplots show the median (line), interquartile range (box) and whiskers extend to 1.5 × the Inter quartile range. Extreme values beyond the whiskers are shown as circles. Ctx – Cortex, Cer – Cerebellum, SN – Substantia nigra. C M - cytoplasmic and mitochondrial fractions extracted from the same sample. Mito/cyto – ratio of mitochondrial Hb compared with cytoplasmic Hb. F-female, M – male. Full densitometry dataset is provided in Supplementary Table 2.

Fig. 3

Scatterplots of mitochondrial/cytoplasmic Hb ratio versus disease duration in years. Scatterplots of the ratio of mitochondrial/cytoplasmic heamoglobin A and B in three tissues. Patient samples are split by sex and a linear model used to fit a trendline. The shaded areas show the 95% confidence limits of the fitted line. Calculations and plotting was conducted using ggplot in R (see supplemental R script which was used for generating figures). Orange circles (female), blue circles (male). Yr – years. Ctx – Cortex, Cer – Cerebellum, SN – Substantia nigra.

Fig. 4

The cellular distribution of HbB appears changed in female cerebellum with disease duration. HbB and COXIV antibodies were utilised to visualise the physical location of mitochondria and HbB in the cerebellum of female individuals with different disease durations. It appears that with longer disease duration the overall quantity of COXIV is reduced and the location of HbB shifts in the longest disease course from in and around Purkinje cells observed in controls, to the granule cell layer in sections from 18 years disease duration. DAPI is used to stain nuclei. Filter sets used are detailed in the methodology.