Can human IgG subclasses distinguish between confirmed and unconfirmed SARS-CoV-2 infections?

Abstract

Background

Immunoglobulin G (IgG) subclasses play a crucial role in the immune response to viral infections. While total IgG levels can generally provide an indication on the immune response, specific IgG subclasses can offer more detailed information about nature of the immune response and stage of the infection. Herein, we addressed the value of both total (t) and SARS-CoV-2-specific (s) IgG-subclasses in distinguishing between infection-confirmed virus-qRT-PCR-positive (IC; V-qRT-PCR-P) and infection-unconfirmed virus-qRT-PCR-unchecked (IU; V-qRT-PCR-UC) Egyptians.

Results

Both the t-IgG2 and 4 means were significantly higher (SH) among the IU subjects, whereas, the s-IgG1 and 3 means were SH among the IC ones. On the gender levels, both the t-IgG2 and 4 means were SH among the IU females, whereas, the mean of the s-IgG1 was SH among the IC females. The t-IgG4 mean was SH among the IU males, whereas, both means of the s-IgG1 and 3 were SH among the IC males. Significant positive correlations (SPC) were recorded between both the t-IgG1 and 3 with the symptom grades (SG) among the IU humans (r² = 0.200 and 0.253, respectively). Also, SPC was noticed between the s-IgG2 and the SG among the IU females (r² = 0.6782). SPC was recorded between both the t-IgG1 and the s-IgG2 with the SG among the IU males (r² = 0.794 and 0.373, respectively). SPC was noticed between the t-IgG3 and the age among the IC males (r² = 0.779).

Conclusion

Although the limitation of the small studied sample size, our results suggest some total and SARS-CoV-2-specific IgG-subclasses as both supplemental and gender-specific immune markers to distinguish between confirmed and unconfirmed SARS-CoV-2 infections.

1. Introduction

Since its emergence in Autumn 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) became wide spread and resulted in global pandemic. As of November, 6, 2023 over 697 million of COVID-19 infections and over 6 million deaths have been reported globally¹. On February 14, 2020 Egypt announced the first case of COVID-19 in Africa². To date 516,023 confirmed infections with 24,613 deaths and 4946 (4.861 per 100000) of new cases have been reported.¹ The most common manifestations of SARS-CoV-2 infection are fever, dry cough, tiredness, dyspnea, myalgia, fatigue and severe respiratory disorders with different degrees of severity3, 4. The fundamental role of the humoral immunity in host response to microbial infection is well defined. Of the induced immunity to SARS and MERS human infections, neutralizing antibodies were the most abundant response.5, 6, 7, 8, 9 SARS-CoV-2 IgM/IgG are highly detectable within the first 3 weeks of disease onset and show significantly higher levels among the patients who develop severe symptoms.¹⁰ Additionally, COVID-19 patients showed elevated levels of IgA, IgM, IgG and/or subclasses to nucleocapsid (N), spike (S) and receptor binding domain (RBD)¹¹. Old males with comorbid diseases showed higher levels of the virus-specific antibodies12, 13 whereas some of asymptomatic and/or recovered individuals showed a low to undetectable IgG levels against virus-N and S proteins.¹⁴ Though, human IgG subclasses exhibit > 90 % sequence identity, yet, they still perform distinct functions. While human IgG2 is a key element against bacterial infection, IgG1 and IgG3 are more dominate against viral infections.¹⁵.

The value of IgG3 was associated with HIV and HCV clearance and neutralization16, 17. Infection with HCV has been associated with beak in IgG1 and IgG3 values and the patients with higher antibody titers were able to clear the virus earlier than the chronically infected ones.¹⁷.

Specific Immunoglobulin isotypes to the N, S, and RBD have been intensively studied among COVID-19 patients¹⁸. RBD-specific IgG1 and IgG3were detected in COVID-19 patients and the higherIgG3 level was connected to viral load.¹⁹ Using a full-length S protein-based flow cytometry assay, all IgG subclasses were detected among COVID-19 patients with a high prevalence of IgG1²⁰. Moreover, high prevalence of IgG1 subclass was also reported to both BRD and N antigens.²¹ The production of IgG subclasses is tightly regulated both by class switching process and secretion of cytokines by the helper and regulatory T-cells.²².

Here we evaluated the value of both the total and SARS-CoV-2-specific IgG subclasses as diagnostic markers to distinguish between infection-confirmed virus-qRT-PCR-positive (IC; V-qRT-PCR-P) and infection-unconfirmed virus-qRT-PCR-unchecked (IU; V-qRT-PCR-UC) Egyptians.

2. Materials and methods

2.1. The studied human cohort and blood samples collection

This study has been conducted on IC (V-qRT-PCR-P) and IU (V-qRT-PCR-UC) Egyptian cohort. The IC subjects were 14, of whom, 8 were females and 6 were males. All the 8 females were symptomatic (S), while, of the 6 males, 5 were S and 1 was asymptomatic (AS). The IU subjects were 25, of whom, 12 were females and 13 were males. Of the 12 females 6 were S and 6 were AS. Of the 13 males, 6 were S and 7 were AS.

Serum samples of the IC humans were kindly provided by Med. Dr. Wael A Hassan (Egypt Center for Research and Regenerative Medicine, Cairo, Egypt). For these subjects, the viral RNA was quantified using the TaqPath COVID-19 CE-IVD RT- PCR Kit (A51738; Thermo Fisher Scientific) according to the manufacturer instructions. Serum samples of the IU humans were collected from both employees and workers at the National Research Centre of Egypt.

The blood collection was carried out in compliance with relevant laws as well as followed the institutional guidelines and the ethical standards of the Declaration of Helsinki. This study has been approved from the Medical Ethical Committee of the National Research Centre (Meeting date: 5.11.2020, Decision number: 20166). All the details about the written informed consent, questionnaire, recorded symptoms and how the number of symptoms was transformed into grade are described in detail in our previously published work.²³.

Sera were separated from the collected blood and being freshly used in the SARS-CoV-2 IgG/IgM AMP rapid test (AMEDA Labordiagnostik GmbH; Graz, Austria) to detect the virus-specific IgM/IgG. This test relies on both the recombinant viral spike 1 (S1) and nucleocapsid (N) proteins as indicated by the manufacturer. The rest of serum aliquots were used to semi-quantitatively detect virus-specific IgM/IgG using the anti-2019 nCoV (N) ELISA kits (WuhanFine Biotech; Cat Nrs. EH4396 and EH4395) according to the manufacturer instructions. Changes of the optical density (OD; absorbance) were measured using a multi-well plate reader (Tecan, Switzerland) at λmax 450 nm and reference wavelength 680 nm.

2.2. Total and virus-specific IgG subclasses detection

Total IgG subclasses were measured using human IgG subclasses ELISA kit (Cat Nrs. 991000; Thermo Fisher scientific) as described by the manufacturer. Specific IgG subclasses were analyzed by combining the above-mentioned anti-2019 nCoV (N) ELISA kit (Wuhan Fine Biotech; Cat Nrs. EH4396 and EH4395) with the human IgG subclasses ELISA kit (Cat Nrs. 991000; Thermo Fisher scientific) as follow: 1) the first kit was used to capture the virus-specific total IgG, 2) the non-specific binding of antibodies was blocked using 5 % fetal bovine serum in the PBS-0.05 % Tween20, 3) the plate wells were then incubated with subclass-specific monoclonal antibodies (Provided in the second kit, 4) the remaining steps were carried out as described by the manufacturer with a minor modification in the incubation time to become one hour in order to optimize the signal to background ratio, 5) and finally the changes in the optical densities were measured using a multi-well plate reader (Tecan, Switzerland) at λmax 450 nm and reference wavelength 680 nm.

2.3. Statistical analysis

Both data analysis and plots were done using the GraphPad PRISM version 5 software. Results were expressed as means ± standard deviations (SD). Significance was calculated by comparing the differences of the means of the studied groups, the student’s t-test was applied, and p-values < 0.05 were considered significant. Correlation analysis was carried out by calculating the square value of the correlation coefficient (r2) for the nonparametric and non-normally distributed data.

3. Results

The average age of the IC humans (n = 14) was 48.15 ranged from10-70 years, whereases, that of the IU subjects (n = 25) was 43.04 ranged from 21-67 years.

Using the rapid test, 13/14 (92.8 %) and 5/25 (20 %) were IgM positive among the IC and IU humans, respectively whereas for the IgG, 14/14 (100 %) and 14/25 (56 %) were positive among the IC and IU humans, respectively. Using the SARS-CoV-2 (N) IgG ELISA kit, 3/14 (21.4 %) and 8/25 (32 %) were IgM positive among the IC and IU humans, respectively, while,14/14 (100 %) and 25/25 (100 %) were IgG positive among the IC and IU humans, respectively.

The overall SARS-CoV-2-(N)-IgG mean value was significantly higher (SH) in the IC than the IU humans (p = 0.000, 0.0013, and 0.0009, respectively (Table1 and Fig. 1). Both t-IgG2 and 4 means were SH in the IU humans than the IC ones (p = 0.0097 and 0.0003, respectively; Table1and Fig. 1). Both the s-IgG1 and the t-IgG3 means were SH in the IC humans than in the IU ones (p = 0.0008 and 00.0012, respectively; Table 2 and Fig. 1).

Table 1.

The overall SARS-CoV-2 (N)-IgG and t-IgG subclasses means in sera from the infection-confirmed virus-qRT-PCR-positive and the infection-unconfirmed virus-qRT-PCR-unchecked subjects.

Measured parameter	SARS-CoV-2 (N)-IgG
Subjects	IC (V-qRT-PCR-P)				IU (V-qRT-PCR-UC)
Mean	0.96				1.51
SD	0.25				0.23
p-value	0.0001 (***)
Measured parameter	t-IgG subclasses
Subclass/subjects	tIgG1/IC	tIgG1/IU	tIgG2/IC	tIgG2/IU	tIgG3/IC	tIgG3/IU	tIgG4/IC	tIgG4/IU
Mean	0.74	0.70	0.55	0.79	1.02	1.09	0.37	0.80
SD	0.25	0.22	0.25	0.21	0.29	0.25	0.11	0.39
p-value	0.5551(NS)		0.0097(**)		0.5251(NS)		0.0003(***)

IC: Infection confirmed; V-qRT-PCR-P: Virus quantitative reverse transcriptase polymerase chain reaction positive; IU: Infection unconfirmed; V-qRT-PCR-UC: Virus quantitative reverse transcription polymerase chain reaction unchecked; t: Total; SD: Standard deviation; NS: Non-significant.

Fig. 1.

Scattered plots showing the SARS-CoV-2 (N)-IgG, total (t) and SARS-CoV-2-specific (s) IgG subclasses in sera from the infection-confirmed (IC; a & c) and the infection-unconfirmed (IU; b & d) humans. Changes in the optical density (OD; absorbance) that reveals antibody binding were recorded using a multi-well plate reader (Tecan, Switzerland) at λmax 450 nm and reference wavelength 680 nm.

Table 2.

The overall SARS-CoV-2 (N)-IgG and SARS-CoV-2-specific IgG subclasses means in the infection-confirmed virus-qRT-PCR-Positive and the infection-unconfirmed virus-qRT-PCR-unchecked humans.

Measured parameter	SARS-CoV-2 (N)-IgG
Subjects	IC (V-qRT-PCR-P)				IU (V-qRT-PCR-UC)
Mean	1.57				1.14
SD	0.23				0.17
p-value	0.0001 (***)
Measured parameter	s-IgG subclasses
Subclass/subjects	tIgG1/IC	tIgG1/IU	tIgG2/IC	tIgG2/IU	tIgG3/IC	tIgG3/IU	tIgG4/IC	tIgG4/IU
Mean	0.37	0.05	0.62	0.55	0.82	0.25	0.09	0.08
SD	0.34	0.03	0.32	0.20	0.56	0.18	0.03	0.04
p-value	0.0008(***)		0.6443(NS)		0.0012(**)		0.0087(**)

IC: Infection confirmed; V-qRT-PCR-P: Virus quantitative reverse transcriptase polymerase chain reaction positive; IU: Infection unconfirmed; V-qRT-PCR-UC: Virus quantitative reverse transcription polymerase chain reaction unchecked; s: Specific: SD: Standard deviation; NS: Non-significant.

On the genders level, the SARS-CoV-2 (N)-IgG means were SH among both genders of the IC humans than the IU ones (p = 0.0013 and 0.0009 for females and males, respectively; Table 3, Fig. 2, Fig. 3). Both the s-IgG1 and 3 means were SH among the IC compared to the IU subjects (p = 0.0008 and 0.0012, respectively Table 2 and Fig. 1). Both the t-IgG2 and 4 means were SH among females of the IU subjects than the IC ones (p = 0.0228 and 0.0205, respectively Table 3 and Fig. 2), whereas, only the mean of the t-IgG4 was SH among males of the IU humans compared to the IC ones (p = 0.0043; Table 3 and Fig. 3).

Table 3.

The SARS-CoV-2 (N)-IgG and t-IgG subclasses means among both genders of the infection-confirmed virus-qRT-PCR-positive and the infection-unconfirmed virus-qRT-PCR-unchecked humans.

Gender	Females
Measured	SARS-CoV-2 (N)-IgG
Subjects	IC (V-qRT-PCR-P)				IU (V-qRT-PCR-UC)
Mean	1.45				0.98
SD	0.21				0.24
p-value	0.0013 (**)
Measured parameter	t-IgG subclasses
Subclass/subjects	tIgG1/IC	tIgG1/IU	tIgG2/IC	tIgG2/IU	tIgG3/IC	tIgG3/IU	tIgG4/IC	tIgG4/IU
Mean	0.71	0.70	0.47	0.80	1.05	1.11	0.38	0.69
SD	0.26	0.24	0.27	0.18	0.33	0.25	0.09	0.28
p-value	0.9692(NS)		0.0228(*)		0.9692(NS)		0.0205(*)
Gender	Males
Measured	SARS-CoV-2 (N)-IgG
Subjects	IC (V-qRT-PCR-P)				IU (V-qRT-PCR-UC)
Mean	1.67				0.94
SD	0.07				0.28
p-value	0.0009 (***)
Measured parameter	t-IgG subclasses
Subclass/subjects	tIgG1/IC	tIgG1/IU	tIgG2/IC	tIgG2/IU	tIgG3/IC	tIgG3/IU	tIgG4/IC	tIgG4/IU
Mean	0.77	0.69	0.65	0.78	0.99	1.07	0.36	0.92
SD	0.25	0.20	0.20	0.25	0.26	0.26	0.15	0.47
p-value	0.5427(NS)		0.2608(NS)		0.5427(NS)		0.0043(**)

IC: Infection confirmed; V-qRT-PCR-P: Virus quantitative reverse transcriptase polymerase chain reaction positive; IU: Infection unconfirmed; V-qRT-PCR-UC: Virus quantitative reverse transcription polymerase chain reaction unchecked; t: Total; SD: Standard deviation; NS: Non-significant.

Fig. 2.

Scattered plots displaying SARS-CoV-2(N)-IgG, total (t) and SARS-CoV-2-specific (s) IgG subclasses among females of the IC (a & c) and the IU (b & d) subjects. Changes in the optical density (OD; absorbance) that reveals antibody binding were recorded using a multi-well plate reader (Tecan, Switzerland) at λmax 450 nm and reference wavelength 680 nm.

Fig. 3.

Scattered plots showing SARS-CoV-2(N)-IgG, total (t) and SARS-CoV-2-specific (s) IgG subclasses levels among males of the IC (a& c) and the IU (b & d) humans. Changes in the optical density (OD; absorbance) that reveals antibody binding were recorded using a multi-well plate reader (Tecan, Switzerland) at λmax 450 nm and reference wavelength 680 nm.

The SARS-CoV-2-specific IgG1mean was SH among females of the IC subjects than in the IU ones (p = 0.0080; Table 4 and Fig. 2) whereas both SARS-CoV-2-specific IgG1 and 3 means were SH among males of the IC humans compared to the IU ones (p = 0.0441 and 0.0043, respectively; Table 4 and Fig. 3).

Table 4.

The SARS-CoV-2 (N)-IgG and s-IgG subclasses means among both genders of the infection-confirmed virus-qRT-PCR-positive and the infection-unconfirmed virus-qRT-PCR-unchecked humans.

Gender	Females
Measured parameter	s-IgG subclasses
Subclass/subjects	sIgG1/IC	sIgG1/IU	sIgG2/IC	sIgG2/IU	sIgG3/IC	sIgG3/IU	sIgG4/IC	sIgG4/IU
Mean	0.32	0.04	0.56	0.52	0.86	0.32	0.08	0.09
SD	0.31	0.01	0.33	0.19	0.71	0.23	0.008	0.06
p-value	0.0080(**)		1.0000(NS)		0.0988(NS)		0.1173(NS)
Gender	Males
Measured parameter	s-IgG subclasses
Subclass/subjects	sIgG1/IC	sIgG1/IU	sIgG2/IC	sIgG2/IU	sIgG3/IC	sIgG3/IU	sIgG4/IC	sIgG4/IU
Mean	0.41	0.06	0.67	0.57	0.78	0.19	0.09	0.06
SD	0.40	0.05	0.35	0.23	0.46	0.10	0.05	0.007
p-value	0.0441(*)		0.8548(NS)		0.0043(**)		0.271(NS)

IC: Infection confirmed; virus quantitative reverse transcriptase polymerase chain reaction positive; IU: Infection unconfirmed; virus quantitative reverse transcription polymerase chain reaction unchecked; s: Specific; SD: Standard deviation; NS: Non-significant.

In the IU subjects, significant positive correlations (SPC) were recorded between both the total IgG1 and 3 and the symptoms grade (SG; r² = 0.200 and 0.253; p = 0.321 and 0.0144, respectively; Table 5). Also, SPC were noticed between the s-IgG2 and the SG among females of the IU humans (r² = 0.6782, p = 0.044; Table 5).

Table 5.

The correlations between both the SARS-CoV-2 (N)-IgG, total and specific IgG subclasses with the symptoms grades and the age of the infection-confirmed virus-qRT-PCR-positive and the infection-unconfirmed virus-qRT-PCR-unchecked humans.

Subjects	Correlated parameters	(r2)	p-value
IC (V-qRT-PCR-P) males	s-IgG2/Symptoms grades	0.794	0.0423 (*)
	Age/Symptoms grades	0.779	0.0474 (*)
IU (V-qRT-PCR-UC)	t-IgG1/Symptoms grades	0.200	0.321 (*)
	t-IgG3/Symptoms grades	0.253	0.0144 (*)
IU (V-qRT-PCR-UC)females	s-IgG2/Symptoms grades	0.6782	0.044 (*)
IU (V-qRT-PCR-UC)males	t-IgG1/Symptoms grades	0.373	0.034 (*)

IC: Infection confirmed; V-qRT-PCR-P: Virus quantitative reverse transcriptase polymerase chain reaction positive; IU: Infection unconfirmed; V-qRT-PCR-UC: Virus quantitative reverse transcription polymerase chain reaction unchecked; t: Total; s: Specific.

A SPC was recorded between the t-IgG1 and the SG among males of the IU humans (r² = 0.373, p = 0.034; Table 5). Similarly, a SPC was noticed between the s-IgG2 and the SG among males of the IC subjects (r² = 0.794, p = 0.0423; Table 5). Likewise, a SPC was recorded between the t-IgG3 and the age among males of the IC subjects (r² = 0.779, p = 0.0474; Table 5).

4. Discussion

The noticed overall SH mean values of the SARS-CoV-2 (N)-IgG among both genders of the IC subjects agree with the fact that this cohort was on high level of disease severity, and therefore were admitted to the hospital. Though IgG4 is the least abundant subclass, accounting for 3–6 % of total IgG in human serum,²⁴ the noticed SH means of both the total IgG2 and 4 among females and particularly the t-IgG4 among males of the IU group suggest the potential impact of these two IgG subclasses as 1) immune markers for the moderate/less severe infections, and/or for the repeated exposure to the SARS-CoV-2 antigens, 2) they could be implicated in the virus clearance (neutralization). The higher level of IgG4 was previously reported both in the late infection and upon repeated exposure to the virus antigens.²⁵ In contrast to our findings, previous studies reported that both IgG2 and IgG4 were less abundant in sera of infected patients with viruses.26, 27 IgG4 is considered as allergen-specific immunotherapy,²⁸ and has been applied to modulate the induced inflammation by parasitosis.²⁹ In the same line, the patients who had deficiency in the IgG4 subclass showed recurrent sinopulmonary infections.³⁰ Additionally, IgG2 is the second most abundant IgG subclass in human serum, accounting for 20–25 % of the total IgG.²⁴ The patients who had deficiency of IgG2 showed severe H1N1 influenza infection.31, 32 .

The observed SH mean of the specific IgG1 among females and the SH means of both the specific IgG1 and 3 among males of the IC humans agree with the previously reported data where both of these IgG subclasses are typical manifestations for type 1 T helper pro-inflammatory response and have fundamental role in protection against viruses.8, 33 Accordingly, measurement of these IgG subclasses might complement and enhance the molecular-based diagnosis of COVID-19.

Matching to our findings, specific IgG1 and IgG3 were the most dominant IgG subclasses among COVID-19 patients.20, 28 It has been also widely documented that the IgG1 and 3 subclasses have higher neutralization capabilities against multiple viruses.34, 16, 17 Of interest, similar IgG subclasses profile was noticed both after natural infection with measles and early upon vaccination where IgG1 was the first seen IgG subclass followed by IgG2, IgG3 and IgG4 which showed gradual increase during the convalescence.³⁵ Moreover, IgG1 and IgG3 are essential for complement proteins fixation and lysis of infected cells through antibody-dependent cellular toxicity.²⁷ Thus, our finding might suggest a possible difference in complement activation between the IC and the IU humans, and this deserves further studies.

Previous study showed induction of higher IgG level with a fucosylated Fc glycans in severe COVID-19 patients which enhanced their interaction with the Fc receptor on monocytes and induced release of inflammatory cytokines leading to severe COVID-19 symptoms.³⁶ In harmony, the recorded SPC between total IgG1 and 3 with the SG of symptomatic IU humans suggest potential implication of these two subclasses with the induced COVID-19 severity which also requires further investigations.

In line with our expectation, a previous study has suggested an association between the SARS-CoV-2 (N)- IgG1 and 3 with COVID-19 severity.¹¹ Of interest, the noticed SPC between the s-IgG2 with the SG among both the IU females, the t-IgG1and s-IgG2 among the IU males suggest potential connection between these IgG subclasses with COVID-19 progression and/or severity.

The recorded SPC between t-IgG3 and age among the IC males suggest this IgG subclass as a gender specific immune marker for SARS-CoV-2 confirmed infection. Previous report revealed that both the virus-specific IgG and IgG subclasses are highly influenced both by age and co-morbidities.¹¹ Of note, both the advanced age and co-morbidities were considered as risk factors and influenced the IgG subclasses levels in relation to the total IgG level in SARS-CoV-2 infection.37, 38 Some reports speculated that the differential profiling of the IgG subclasses might enable differentiation between confirmed and unconfirmed SARS-CoV-2 infections, could demonstrate a connection to disease severity grade, indicate the nature of induced immunity and could define the repeated exposure to the virus antigens.8, 34

5. Conclusion

Although limitation of the small studied sample size, our findings suggest some total and specific IgG subclasses as both supplemental and gender-specific immune markers to distinguish between confirmed and unconfirmed SARS-CoV-2 infections and reveal levels of disease severity.

6. Authors’ contributions

MB made a substantial contribution to the conception and design of the work, providing the funding to the whole work and precisely revised the manuscript.

MHN, DA and RN did the experimental work.

MHN performed the data analysis and interpretation and prepared the manuscript draft.

All authors read and approved the final manuscript.

7. Consent for publication

Not applicable.

8 Ethics approval and consent to participate.

This study was approved by the Medical Ethical Committee of the National Research Centre (Meeting date: 5.11.2020, Decision number: 20166).

8. Availability of data and materials

The corresponding author welcome to deliver the row data upon request.

Funding

This study was funded by the mandatory grant MP120803 from the National Research Centre of Egypt provided to Mahmoud Mohamed Bahgat.

CRediT authorship contribution statement

Mahmoud Mohamed Bahgat: Funding acquisition, Study conceptualization, Writing, Reviewig, Editing, Discussing the results and finalizing the manuscript. Mohamed Hassan Nasraa: Methodology, Investigation, Data analysis analysis, Discussing results, Writing the original paper draft and performing the changes and corrections. Rola Nadeem: Methodology, Investigation and Discussing results. Khaled Amer: Providing well-characterized samples and discussing the results. Wael A. Hassan: Providing well-characterized samples and discussing the results. Ahmed Abd EL-Raouf: Providing well-characterized samples and discussing the results. Dina Nadeem Abd-Elshafy: Methodology, Investigation and Discussing results.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.