Relationship between plasma proteome and hydrocephalus: a two-sample Mendelian randomization analysis

doi:10.21203/rs.3.rs-4982718/v1

Background: Hydrocephalus is a serious condition that affects individuals across all age groups, arising from various etiologies. The substantial efficacy of plasma proteome in numerous diseases has been substantiated by multiple studies. We conducted a two-sample Mendelian randomization analysis to investigate the potential association between plasma proteome and hydrocephalus.

Methods: By conducting a two-sample Mendelian randomization analysis, we identified 98 plasma proteins that either increased or decreased the risk of communicating hydrocephalus out of a total of 4,907 proteins. The inverse-variance weighted method for causal effect estimation, as well as the weighted median, maximum likelihood, and MR Egger regression methods, were further employed for sensitivity analyses after selecting the top 20 proteins. Additionally, gene enrichment analysis was performed to uncover relevant functional pathways.

Results: Elevated levels of ADGRF1 (OR=2.728, 95% CI [1.266-5.879], P=0.010), APP (OR=2.923, 95% CI [1.328-6.431], P=0.008), DUSP13 (OR=3.201, 95% CI [1.062-9.647], P=0.039), EPHA3 (OR=4.341, 95% CI [1.031-18.288], P=0.045), HSPA1B (OR=4.578, 95% CI [1.431-14.643], P=0.010), HYAL1 (OR=3.075, 95% CI [1.375-6.881], P=0.006), KIRREL3 (OR=6.067, 95% CI [1.278-28.806], P=0.023), RAD23B (OR=6.825, 95% CI [1.245-37.415], P=0.027), SERPINE1 (OR=3.768, 95% CI [1.452-9.777], P=0.006), SNPH (OR=4.019, 95% CI [1.059-15.252], P=0.041), SRP14 (OR=6.292, 95% CI [1.193-33.189], P=0.030), and TMEFF1 (OR=2.830, 95% CI [1.097-7.300], P=0.031) were associated with an increased risk of hydrocephalus. Additionally, these relevant proteins primarily participate in biological processes associated with axon development, axonogenesis, axon guidance, and neuron projection guidance.

Conclusion: Higher genetically predicted levels of protein ADGRF1, APP, DUSP13, EPHA3, HSPA1B, HYAL1, KIRREL3, RAD23B, SERPINE1, SNPH, SRP14 and TMEFF1 are associated with an increased risk of hydrocephalus. The findings provide valuable insights into potential therapeutic targets, protective mechanisms, and underlying biological processes, ultimately leading to the development of more effective and personalized treatment strategies for individuals with hydrocephalus.

plasma protein

communicating hydrocephalus

Mendelian randomization

Hydrocephalus is a severe condition that can impact individuals across all age groups, arising from various origins. Although the causes of hydrocephalus are diverse, the acute and chronic symptoms associated with the condition tend to be common among patients[1]. Hydrocephalus can pose an immediate threat to life and significantly impact various other critical neurological comorbidities and conditions[2]. From a neurosurgical standpoint, treating hydrocephalus is considered the foundational or starting point in the management of spina bifida[3, 4]. In cases where unexplained hydrocephalus or ventriculomegaly might be attributed to the loss of brain volume rather than increased intracranial pressure, it's important to consider a deficiency in methylenetetrahydrofolate reductase. Early intervention in such cases could lead to improved outcomes[5]. The origins of hydrocephalus are varied, often stemming from pathophysiological mechanisms that disrupt CSF flow. Various classifications of hydrocephalus have been proposed by numerous researchers, resulting in distinct categories such as communicating hydrocephalus and non-communicating hydrocephalus[6]. The most frequent treatment for excessive CSF involves surgically placing a catheter to divert the fluid to another body cavity. Yet, this procedure only mitigates 10–20% of the neurological impairments in newborns with hydrocephalus that begins in utero, indicating that hydrocephalus may not solely be a disorder related to CSF dynamics but also a condition affecting the brain itself[7]. Although blockage of cerebrospinal fluid is the most common cause of congenital hydrocephalus, the underlying cellular and molecular mechanisms are not yet well understood[8].

Proteins found in blood plasma have a long history of use as biomarkers for diseases, such as alpha-fetoprotein (AFP), CA-125, carcinoembryonic antigen (CEA), CA15-3, and CA19-9[9–11]. At this juncture, it is reasonable to question whether all protein biomarkers for hydrocephalus have been identified and whether the current protein detection technologies meet the requirements for cerebrospinal fluid liquid biopsy.

In our traditional analyses, proving causal relationships between diseases and other factors is challenging. Mendelian randomization offers an alternative approach for investigating causality in epidemiological studies by utilizing hypothetical genetic variants to satisfy the assumptions of an instrumental variable[12]. Mendelian randomization employs instrumental variable analysis to assess causal hypotheses using non-experimental data. It commonly utilizes single nucleotide polymorphisms as instrumental variables for assumed risk factors. Additionally, this method lowers the likelihood of reverse causation bias because genetic variations are fixed and remain unaffected by the state of the disease[13].

In this study, Mendelian randomization analysis was conducted using 4,903 plasma proteins as exposure factors and hydrocephalus as the outcome. This approach effectively identified plasma proteins that either increase the risk or offer protection against hydrocephalus. This has significant implications for the treatment of hydrocephalus, as identifying these proteins could lead to the development of targeted therapies. By understanding the genetic basis of protein interactions with hydrocephalus, researchers can explore new preventive strategies and therapeutic interventions, potentially improving patient outcomes in cases of this condition.

Study design

To investigate the potential causal link between plasma proteins and communicating hydrocephalus, we conducted a two-sample Mendelian randomization (MR) analysis. Single nucleotide polymorphisms (SNPs) were used as instrumental variables (IVs) in this study. By employing SNPs in a manner analogous to randomized controlled trials, this modeling approach enables the identification of causal relationships between exposure factors, such as plasma proteins, and outcome variables, like hydrocephalus. The current study employed data from previously published research that had been approved by institutional review boards and obtained informed consent from the original study participants, negating the need for additional sanctions.

Data sources

We gathered GWAS data for 4,907 plasma proteins in the Icelandic population from the website https://www.decode.com/summarydata/. And we obtained the GWAS summary statistics for communicating hydrocephalus from the FinnGen database release 9 (R9), available at https://storage.googleapis.com/finngen-public-data-r9/summary_stats/finngen_R9_G6_HCCOMM.gz.

The hydrocephalus dataset consisted of 488 individuals diagnosed with the condition and 375,610 healthy controls. The selection of SNPs adhered to a stringent set of criteria. The first criterion involved the inclusion of SNPs that showed a strong association with plasma proteins at a genome-wide significance level (p = 5×10E-8). To ensure the independence of the selected SNPs and minimize the impact of linkage disequilibrium (LD) on the results, a rigorous threshold of r²= 0.001 was applied within a window size of 10,000 kb. This approach effectively reduced any potential bias arising from LD. Moreover, the F-statistic of the SNPs was used to evaluate the correlation between the instrumental variables (IVs) and the exposure factors. As a general principle, IVs with an F-statistic exceeding 10 are considered to be unbiased indicators.

Statistical analysis

To ensure a valid interpretation of the MR analysis results, it is crucial to consider the following hypotheses: (i) The association between the instrumental variables (IVs) and plasma proteins has been robustly established in previous studies; (ii) The influence of IVs on hydrocephalus is solely mediated through their effects on plasma protein levels; (iii) The IVs confirmed that the relationship between plasma proteins and hydrocephalus is free from confounding factors.

Genetic variation may introduce bias in the causal estimates via a single pathway, known as horizontal pleiotropy, which violates the assumptions of Mendelian randomization (MR). To address this issue, the MR analysis employed three distinct analytical methods, each based on a different model of horizontal pleiotropy. Comparing the results from these approaches strengthens the credibility of the findings by demonstrating consistency across methods. The primary analysis employed an inverse variance-weighting (IVW) approach, which yields the most accurate estimates under the assumption that all SNPs serve as valid instrumental variables (IVs). In cases where any SNP does not meet the IV criteria, a random-effects IVW method is applied to provide an adjusted estimate that factors in the standard error of each ratio and potential heterogeneity. Additionally, the weighted median method was used, requiring that a minimum of 50% of the SNPs involved are valid IVs. This approach ranks the SNPs by their weights and calculates the median from the associated distribution function based on experimental outcomes. Furthermore, MR-Egger regression allows for an estimate of the effect if the genetic instrument remains free from pleiotropic influences. Pleiotropic effects were assessed using the intercept from the MR-Egger regression; an intercept close to zero indicates that no directional multiplicative effect is evident. To assess the robustness of the results, a comprehensive "leave-one-out" analysis was conducted[14]. This method involved iteratively excluding one single-nucleotide polymorphism (SNP) at a time to determine its impact on the overall findings.

All statistical computations were performed using the TwoSampleMR package (version 0.5.5) within the R software environment (version 4.0.3). The application of these stringent analytical techniques aimed to ensure the reliability and validity of the study's conclusions.

Functional Enrichment Analysis

We cross-referenced the differentially expressed genes with the significant genes identified through Mendelian randomization analysis to obtain a set of plasma protein-related genes. These genes were then subjected to enrichment analysis. Using the R statistical software and packages such as 'clusterProfiler', 'org.Hs.eg.db', 'enrichplot', 'ggplot2', and 'GOplot', we conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) biological process enrichment analyses for the identified genes. An adjusted p-value threshold of < 0.05 was used to determine statistical significance.

To investigate the protein-protein interaction relationships of the key proteins identified through Mendelian randomization analysis, we utilized the online tool GeneMANIA, accessible at http://genemania.org/.

Screening of key plasma proteins

By conducting a Mendelian randomization analysis with 4,907 plasma proteins as the exposure and hydrocephalus as the outcome, we identified a total of 98 plasma proteins that exhibited either a risk-increasing or protective effect on hydrocephalus. These 98 plasma proteins were then subjected to further critical analyses(Fig. 1).

The causal relationship between differentially plasma protein and hydrocephalus

From the 98 key plasma proteins, we selected the 20 most significant proteins that had the greatest impact on hydrocephalus for further analysis(the MR results are shown in Table 1 and supplementary materials). In this study, we identified 80 single nucleotide polymorphisms (SNPs) strongly associated with hydrocephalus. The results of the Inverse Variance Weighting (IVW) analysis indicated the following associations: Elevated levels of ADGRF1 (OR = 2.728, 95% CI [1.266–5.879], P = 0.010), APP (OR = 2.923, 95% CI [1.328–6.431], P = 0.008), DUSP13 (OR = 3.201, 95% CI [1.062–9.647], P = 0.039), EPHA3 (OR = 4.341, 95% CI [1.031–18.288], P = 0.045), HSPA1B (OR = 4.578, 95% CI [1.431–14.643], P = 0.010), HYAL1 (OR = 3.075, 95% CI [1.375–6.881], P = 0.006), KIRREL3 (OR = 6.067, 95% CI [1.278–28.806], P = 0.023), RAD23B (OR = 6.825, 95% CI [1.245–37.415], P = 0.027), SERPINE1 (OR = 3.768, 95% CI [1.452–9.777], P = 0.006), SNPH (OR = 4.019, 95% CI [1.059–15.252], P = 0.041), SRP14 (OR = 6.292, 95% CI [1.193–33.189], P = 0.030), and TMEFF1 (OR = 2.830, 95% CI [1.097-7.300], P = 0.031) were associated with an increased risk of hydrocephalus. Conversely, higher levels of ADI1 (OR = 0.301, 95% CI [0.112–0.807], P = 0.017), BIN1 (OR = 0.332, 95% CI [0.112–0.980], P = 0.046), FABP4 (OR = 0.295, 95% CI [0.099–0.878], P = 0.028), GPI (OR = 0.164, 95% CI [0.043–0.629], P = 0.008), IGFBP5 (OR = 0.259, 95% CI [0.078–0.857], P = 0.027), SDCBP2 (OR = 0.316, 95% CI [0.118–0.843], P = 0.021), SLC5A5 (OR = 0.234, 95% CI [0.077–0.714], P = 0.011), and SRA1 (OR = 0.360, 95% CI [0.140–0.925], P = 0.034) were associated with a decreased risk of hydrocephalus(Fig. 2).

Table 1

The MR results of the relationship between 20 key plasma proteins and communicating hydrocephalus.
Exposures	Methods	SNP	Beta	SE	OR	OR_lci95	OR_uci95	P-value
ADGRF1	MR Egger	6	1.451	1.329	4.269	0.315	57.787	0.336
	Weighted median	6	0.762	0.398	2.142	0.983	4.669	0.055
	Inverse variance weighted	6	1.004	0.392	2.728	1.266	5.879	0.010
	Simple mode	6	0.360	0.674	1.434	0.382	5.375	0.616
	Weighted mode	6	0.733	0.468	2.081	0.832	5.205	0.178
ADI1	MR Egger	3	-2.487	1.627	0.083	0.003	2.015	0.369
	Weighted median	3	-1.374	0.526	0.253	0.090	0.710	0.009
	Inverse variance weighted	3	-1.199	0.503	0.301	0.112	0.807	0.017
	Simple mode	3	-1.413	0.685	0.243	0.064	0.932	0.175
	Weighted mode	3	-1.392	0.612	0.249	0.075	0.825	0.151
APP	MR Egger	4	0.802	1.034	2.229	0.294	16.922	0.519
	Weighted median	4	1.117	0.485	3.055	1.180	7.911	0.021
	Inverse variance weighted	4	1.072	0.402	2.923	1.328	6.431	0.008
	Simple mode	4	1.660	0.762	5.260	1.180	23.440	0.118
	Weighted mode	4	1.112	0.555	3.039	1.024	9.022	0.139
BIN1	MR Egger	4	-1.462	1.411	0.232	0.015	3.684	0.409
	Weighted median	4	-0.977	0.476	0.376	0.148	0.957	0.040
	Inverse variance weighted	4	-1.103	0.553	0.332	0.112	0.980	0.046
	Simple mode	4	-0.936	0.674	0.392	0.105	1.470	0.259
	Weighted mode	4	-0.954	0.496	0.385	0.146	1.019	0.150
DUSP13	MR Egger	3	0.526	4.059	1.693	0.001	4830.730	0.918
	Weighted median	3	1.104	0.640	3.016	0.859	10.582	0.085
	Inverse variance weighted	3	1.163	0.563	3.201	1.062	9.647	0.039
	Simple mode	3	1.097	0.757	2.994	0.679	13.212	0.285
	Weighted mode	3	1.092	0.780	2.980	0.646	13.752	0.297
EPHA3	MR Egger	3	2.697	1.751	14.838	0.480	458.805	0.367
	Weighted median	3	1.568	0.834	4.799	0.936	24.600	0.060
	Inverse variance weighted	3	1.468	0.734	4.341	1.031	18.288	0.045
	Simple mode	3	1.576	1.099	4.838	0.561	41.724	0.288
	Weighted mode	3	1.723	1.021	5.603	0.758	41.424	0.233
FABP4	MR Egger	3	-1.007	2.143	0.365	0.005	24.382	0.720
	Weighted median	3	-1.369	0.621	0.254	0.075	0.858	0.027
	Inverse variance weighted	3	-1.220	0.556	0.295	0.099	0.878	0.028
	Simple mode	3	-1.457	0.797	0.233	0.049	1.110	0.209
	Weighted mode	3	-1.405	0.700	0.245	0.062	0.968	0.183
GPI	MR Egger	4	-3.763	1.366	0.023	0.002	0.338	0.110
	Weighted median	4	-2.483	0.906	0.083	0.014	0.492	0.006
	Inverse variance weighted	4	-1.805	0.685	0.164	0.043	0.629	0.008
	Simple mode	4	-2.389	1.157	0.092	0.009	0.887	0.131
	Weighted mode	4	-2.427	0.980	0.088	0.013	0.603	0.090
HSPA1B	MR Egger	5	1.880	3.494	6.555	0.007	6178.464	0.628
	Weighted median	5	1.807	0.755	6.094	1.388	26.754	0.017
	Inverse variance weighted	5	1.521	0.593	4.578	1.431	14.643	0.010
	Simple mode	5	2.036	1.098	7.659	0.891	65.866	0.137
	Weighted mode	5	1.987	1.113	7.296	0.823	64.696	0.149
HYAL1	MR Egger	5	0.290	0.780	1.337	0.290	6.170	0.735
	Weighted median	5	1.102	0.484	3.011	1.166	7.770	0.023
	Inverse variance weighted	5	1.123	0.411	3.075	1.375	6.881	0.006
	Simple mode	5	1.920	0.747	6.819	1.578	29.459	0.062
	Weighted mode	5	0.718	0.522	2.050	0.737	5.698	0.241
IGFBP5	MR Egger	4	-1.531	1.675	0.216	0.008	5.762	0.457
	Weighted median	4	-1.469	0.707	0.230	0.058	0.920	0.038
	Inverse variance weighted	4	-1.350	0.610	0.259	0.078	0.857	0.027
	Simple mode	4	-1.511	0.912	0.221	0.037	1.318	0.196
	Weighted mode	4	-1.528	0.825	0.217	0.043	1.093	0.161
KIRREL3	MR Egger	3	6.283	3.188	535.603	1.036	277005.049	0.299
	Weighted median	3	1.102	0.844	3.010	0.575	15.755	0.192
	Inverse variance weighted	3	1.803	0.795	6.067	1.278	28.806	0.023
	Simple mode	3	1.100	1.177	3.004	0.299	30.172	0.449
	Weighted mode	3	1.100	1.042	3.004	0.390	23.143	0.402
RAD23B	MR Egger	3	12.999	7.485	442061.940	0.188	1039988046972.450	0.333
	Weighted median	3	1.524	1.046	4.590	0.591	35.657	0.145
	Inverse variance weighted	3	1.921	0.868	6.825	1.245	37.415	0.027
	Simple mode	3	1.046	1.337	2.847	0.207	39.127	0.516
	Weighted mode	3	1.131	1.315	3.099	0.235	40.798	0.480
SDCBP2	MR Egger	6	-1.003	1.554	0.367	0.017	7.717	0.554
	Weighted median	6	-1.084	0.589	0.338	0.107	1.073	0.066
	Inverse variance weighted	6	-1.153	0.501	0.316	0.118	0.843	0.021
	Simple mode	6	-1.127	0.808	0.324	0.066	1.580	0.222
	Weighted mode	6	-1.097	0.767	0.334	0.074	1.501	0.212
SERPINE1	MR Egger	6	2.978	1.428	19.658	1.197	322.879	0.105
	Weighted median	6	0.771	0.644	2.163	0.612	7.643	0.231
	Inverse variance weighted	6	1.327	0.486	3.768	1.452	9.777	0.006
	Simple mode	6	0.466	1.125	1.594	0.176	14.445	0.696
	Weighted mode	6	0.394	0.942	1.483	0.234	9.401	0.693
SLC5A5	MR Egger	3	-0.304	1.773	0.738	0.023	23.831	0.892
	Weighted median	3	-1.265	0.625	0.282	0.083	0.961	0.043
	Inverse variance weighted	3	-1.451	0.568	0.234	0.077	0.714	0.011
	Simple mode	3	-1.282	0.852	0.278	0.052	1.475	0.272
	Weighted mode	3	-1.223	0.681	0.294	0.077	1.117	0.214
SNPH	MR Egger	5	0.235	2.600	1.265	0.008	206.621	0.934
	Weighted median	5	1.369	0.796	3.933	0.827	18.709	0.085
	Inverse variance weighted	5	1.391	0.680	4.019	1.059	15.252	0.041
	Simple mode	5	2.470	1.240	11.820	1.040	134.367	0.117
	Weighted mode	5	1.864	1.176	6.450	0.643	64.689	0.188
SRA1	MR Egger	3	-2.332	2.219	0.097	0.001	7.523	0.484
	Weighted median	3	-1.002	0.501	0.367	0.137	0.980	0.045
	Inverse variance weighted	3	-1.022	0.482	0.360	0.140	0.925	0.034
	Simple mode	3	-0.985	0.653	0.373	0.104	1.342	0.270
	Weighted mode	3	-1.028	0.560	0.358	0.119	1.071	0.208
SRP14	MR Egger	3	2.359	2.735	10.581	0.050	2251.952	0.547
	Weighted median	3	1.290	1.000	3.633	0.512	25.778	0.197
	Inverse variance weighted	3	1.839	0.848	6.292	1.193	33.189	0.030
	Simple mode	3	1.006	1.175	2.734	0.273	27.373	0.482
	Weighted mode	3	1.074	1.202	2.928	0.277	30.892	0.466
TMEFF1	MR Egger	4	2.911	1.813	18.384	0.526	642.062	0.250
	Weighted median	4	1.217	0.549	3.375	1.150	9.903	0.027
	Inverse variance weighted	4	1.040	0.483	2.830	1.097	7.300	0.031
	Simple mode	4	1.074	0.789	2.927	0.623	13.753	0.267
	Weighted mode	4	1.302	0.684	3.677	0.962	14.056	0.153
Figue3. Key gene enrichment analysis(A: Gene Ontology (GO) enrichment analysis; B: Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis )

Key gene enrichment analysis

Gene Ontology (GO) enrichment analysis of the key genes showed that they are mainly involved in biological processes related to axon development, axonogenesis, axon guidance, and neuron projection guidance. Additionally, the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that these genes are significantly enriched in pathways such as Influenza A, Cell adhesion molecules, and Cytokine − cytokine receptor interaction(Fig. 3A-B). To investigate the relationships among these 98 plasma proteins, we have illustrated the regulatory network of interactions between them(Fig. 4).

Communicative hydrocephalus occurs when there is unimpeded flow of cerebrospinal fluid (CSF), resulting from hindered absorption of CSF through the usual drainage pathways or excessive accumulation of CSF due to increased production. The pathological progression of this condition involves a disturbance in water balance within the ventricular system. Although many molecular aspects regarding fluid exchange during hydrocephalus still need clarification, existing literature suggests that specific channel proteins like AQP4 play a vital role in maintaining brain water equilibrium. Previous studies have also shown that the production and composition of CSF are tightly regulated by specific channels, transporters, and pumps that modify plasma filtrate to generate CSF. Additionally, several investigations have emphasized the crucial pathological processes underlying the development of hydrocephalus, which include harm to endothelial cells of the blood-brain barrier (BBB), as well as convective mechanisms[15–16].

Plasma proteins play an important role in maintaining the water balance of the internal environment and regulating various functional proteins. Recently, the advancement of plasma proteomics has led to the identification of an increasing number of proteins offering valuable insights into an individual's health status and hold great potential for revolutionizing disease diagnosis and therapeutic monitoring. However, the specific role of plasma proteins in the exchange process between cerebrospinal fluid and blood remains unexplored. And the exact correlation between hydrocephalus and plasma proteins has yet to be fully elucidated[17–19].

Through the identification of crucial plasma proteins that demonstrate either a propensity to increase the risk or provide protection against hydrocephalus, this study sets the stage for further investigation into the complex and perplexing causes and development of communicating hydrocephalus. Recent genetic association analyses of the plasma proteome showed great casual relation to various diseases except the communicating hydrocephalus. Therefore, the proteins found to be associated with an increased risk of hydrocephalus, such as ADGRF1, APP, and SERPINE1, could serve as potential targets for drug development. Conversely, the proteins associated with a decreased risk, like ADI1, FABP4, and IGFBP5, may provide insights into protective mechanisms that could be harnessed for preventive or therapeutic interventions. Moreover, the enrichment analyses highlighting the involvement of these key genes in axon development, axonogenesis, and neuron projection guidance suggest that targeting these biological processes could be a promising approach for treating hydrocephalus. The identification of relevant pathways, such as cell adhesion molecules and cytokine − cytokine receptor interaction, further expands our understanding of the underlying mechanisms and offers additional avenues for therapeutic exploration.

By elucidating the intricate causal associations between plasma proteins and hydrocephalus, the present study establishes a foundation for promisingly advancing our understanding of the pathomechanism underlying communicating hydrocephalus, encompassing plasma filtration across the blood-brain barrier and the mechanisms governing cerebrospinal fluid distribution. Clinicians may also be able to use this information to stratify patients based on their plasma protein profiles, enabling tailored treatment plans that optimize outcomes and minimize adverse effects. Furthermore, the insights gained from this research could facilitate the discovery of biomarkers for early detection and monitoring of hydrocephalus[20], allowing for timely interventions and improved patient care.

However, it is important to acknowledge the limitations of our research. Firstly, the hydrocephalus dataset was sourced from FinnGen, which represents a relatively homogeneous population. Consequently, caution should be exercised when generalizing our findings to other populations in order to ensure their reliability. Furthermore, while our study supports the MR hypothesis, there is no justification for assuming any significant instrumental bias. Lastly, although we conducted differential gene enrichment analysis and confirmed the differential expression of related functional pathways, the lack of experimental verification has impacted the robustness of our study results.

The Mendelian randomization analysis of plasma proteins and hydrocephalus identified risk-related plasma proteins, which represented a significant step forward in our understanding of communicating hydrocephalus. The findings offer valuable insights into potential therapeutic targets, protective mechanisms, and underlying biological processes, which could ultimately lead to the development of more effective and personalized treatment strategies for individuals affected by hydrocephalus.

Conflict of interest:

The authors declare no conflict of interest in this study.

Funding:

Jiangxi Provincial Health Commission Science and Technology Plan (grant number: 202130441) and Science and Technology Research Project of Education Department of Jiangxi province(grant number: GJJ220C231).

Author Contribution

SD: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Visualization, Writing–original draft, Writing–review and editing. JT: Conceptualization, Investigation, Data curation, Writing–original draft. YD: Data curation, Investigation, Writing–review and editing. MW: Project administration, Resources, Supervision, Validation, Writing–review and editing. HF: Funding acquisition, Project administration, Resources, Supervision, Validation, Writing–review and editing.

Acknowledgments:

We express our gratitude to the authors and researchers of all genome-wide association studies (GWAS) from which we utilized summary statistics datasets. We extend special appreciation to the researchers and administrators of the IEU GWAS database project for their efforts in making the GWAS data publicly accessible, as well as to the participants and researchers involved in the FinnGen study for providing us with the GWAS dataset pertaining to communicating hydrocephalus.

Data Availability

The GWAS data for plasma proteins are available from the website at https://www.decode.com/summarydata/. And the GWAS summary statistics for communicating hydrocephalus from the FinnGen database release 9 (R9), available at https://storage.googleapis.com/finngen-public-data-r9/summary_stats/finngen_R9_G6_HCCOMM.gz.

Hochstetler A, Raskin J, Blazer-Yost BL: Hydrocephalus: historical analysis and considerations for treatment. Eur J Med Res 2022, 27:168.
Ware AL, Kulesz PA, Orkisz JS, Arrington CN, Bowman RM, Fletcher JM: Long-Term Intellectual and Fine Motor Outcomes in Spina Bifida Are Related to Myelomeningocele Repair and Shunt Intervention History. J Int Neuropsychol Soc 2020, 26:364–371.
Fieggen G, Fieggen K, Stewart C, Padayachy L, Lazarus J, Donald K, Dix-Peek S, Toefy Z, Figaji A: Spina bifida: a multidisciplinary perspective on a many-faceted condition. S Afr Med J 2014, 104:213–217.
McCarthy DJ, Sheinberg DL, Luther E, McCrea HJ: Myelomeningocele-associated hydrocephalus: nationwide analysis and systematic review. Neurosurg Focus 2019, 47:E5.
Aljassim N, Alfadhel M, Nashabat M, Eyaid W: Clinical presentation of seven patients with Methylenetetrahydrofolate reductase deficiency. Mol Genet Metab Rep 2020, 25:100644.
James AE, Jr., Flor WJ, Novak GR, Strecker EP, Burns B, Epstein M: Experimental hydrocephalus. Exp Eye Res 1977, 25 Suppl:435–459.
Guerra MM, Henzi R, Ortloff A, Lichtin N, Vio K, Jimenez AJ, Dominguez-Pinos MD, Gonzalez C, Jara MC, Hinostroza F, et al.: Cell Junction Pathology of Neural Stem Cells Is Associated With Ventricular Zone Disruption, Hydrocephalus, and Abnormal Neurogenesis. J Neuropathol Exp Neurol 2015, 74:653–671.
Tully HM, Dobyns WB: Infantile hydrocephalus: a review of epidemiology, classification and causes. Eur J Med Genet 2014, 57:359–368.
Bhardwaj M, Gies A, Werner S, Schrotz-King P, Brenner H: Blood-Based Protein Signatures for Early Detection of Colorectal Cancer: A Systematic Review. Clin Transl Gastroenterol 2017, 8:e128.
Duffy MJ, McDermott EW, Crown J: Blood-based biomarkers in breast cancer: From proteins to circulating tumor cells to circulating tumor DNA. Tumour Biol 2018, 40:1010428318776169.
Fuzery AK, Levin J, Chan MM, Chan DW: Translation of proteomic biomarkers into FDA approved cancer diagnostics: issues and challenges. Clin Proteomics 2013, 10:13.
Bowden J, Holmes MV: Meta-analysis and Mendelian randomization: A review. Res Synth Methods 2019, 10:486–496.
Larsson SC, Butterworth AS, Burgess S: Mendelian randomization for cardiovascular diseases: principles and applications. Eur Heart J 2023, 44:4913–4924.
Xu D, Wang X, Zhou D: Exploring the Potential Association between COVID-19 and Thyroid Cancer: A Mendelian Randomization Study. ACS Omega 2023, 8:49158–49164.
Xu H. New concept of the pathogenesis and therapeutic orientation of acquired communicating hydrocephalus. Neurol Sci. 2016;37(9):1387–1391. doi:10.1007/s10072-016-2589-7
Geyer PE, Kulak NA, Pichler G, Holdt LM, Teupser D, Mann M. Plasma Proteome Profiling to Assess Human Health and Disease. Cell Syst. 2016;2(3):185–195. doi:10.1016/j.cels.2016.02.015
Yang C, Farias FHG, Ibanez L, et al. Genomic atlas of the proteome from brain, CSF and plasma prioritizes proteins implicated in neurological disorders. Nat Neurosci. 2021;24(9):1302–1312. doi:10.1038/s41593-021-00886-6
Geyer PE, Voytik E, Treit PV, et al. Plasma Proteome Profiling to detect and avoid sample-related biases in biomarker studies. EMBO Mol Med. 2019;11(11):e10427. doi:10.15252/emmm.201910427
Desai B, Hsu Y, Schneller B, Hobbs JG, Mehta AI, Linninger A. Hydrocephalus: the role of cerebral aquaporin-4 channels and computational modeling considerations of cerebrospinal fluid. Neurosurg Focus. 2016;41(3):E8. doi:10.3171/2016.7.FOCUS16191
Huang R, Chen Z, He L, He N, Xi Z, Li Z, Deng Y, Zeng X: Mass spectrometry-assisted gel-based proteomics in cancer biomarker discovery: approaches and application. Theranostics 2017, 7:3559–3572.

No competing interests reported.

Relationship between plasma proteome and hydrocephalus: a two-sample Mendelian randomization analysis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study design

Data sources

Statistical analysis

Functional Enrichment Analysis

Results

Screening of key plasma proteins

The causal relationship between differentially plasma protein and hydrocephalus

Key gene enrichment analysis

Discussion

Conclusion

Declarations