Comparison of the oxidative potential of primary (POA) and secondary (SOA) organic aerosols derived from α-pinene and gasoline engine exhaust precursors

Background: Primary (POA) and secondary (SOA) organic aerosols, deriving from both anthropogenic and biogenic sources, represent a major fraction of ambient particulate matter (PM) and play an important role in the etiology of respiratory and cardiovascular diseases, largely through systemic inflammation and cellular oxidative stress. The relative contributions of these species to the inhalation burden, however, are rather poorly characterized. In this study, we measured the in vitro oxidative stress response of alveolar macrophages exposed to primary and secondary PM derived from both anthropogenic and biogenic sources. Methods: POA and SOA were generated within an oxidation flow reactor (OFR) fed by pure, aerosolized α-pinene or gasoline engine exhaust, as representative emissions of biogenic and anthropogenic sources, respectively. The OFR utilized an ultraviolet (UV) lamp to achieve an equivalent atmospheric aging process of several days. Results: Anthropogenic SOA produced the greatest oxidative response (1900 ± 255 µg-Zymosan/mg-PM), followed by biogenic (α-pinene) SOA (1321 ± 542 µg-Zymosan/mg-PM), while anthropogenic POA produced the smallest response (51.4 ± 64.3 µg-Zymosan/mg-PM). Conclusions: These findings emphasize the importance of monitoring and controlling anthropogenic emissions in the urban atmosphere, while also taking into consideration spatial and seasonal differences in SOA composition. Local concentrations of biogenic and anthropogenic species contributing to the oxidative potential of ambient PM may vary widely, depending on the given region and time of year, due to factors such as surrounding vegetation, proximity to urban areas, and hours of daylight.


Amendments from Version 1
Five revisions were made to the manuscript based on referee Ning Zhi's comments: 1. Text was added to support the assumption that hydroxyl radicals were the major products of H2O oxidation in the OFR (response to comment 1) 2. Text was added to clarify experimental conditions during sample collection in the SOA condition (response to comment 2) 3. Text was added to specify details of macrophage cellline proliferation, harvesting, and preparation for PM treatments (response to comment 4) 4. Text was added to specify details of PM sample preparation and treatment conditions in the macrophage assay (response to comment 5) 5. Text was added to address possible role of PM species water-solubility on cellular oxidative stress response (response to comment 6)

Introduction
A large fraction of ambient particulate matter (PM) in the urban atmosphere consists of a mixture of primary organic aerosols (POA), derived from anthropogenic and biogenic PM sources, as well as secondary organic aerosols (SOA) produced during the photo-oxidation of both types of POA (Baltensperger et al., 2005;Després et al., 2012). Urban PM can consist of up to 90% SOA, the majority originating from primary biogenic aerosols, including the monoterpene α-pinene, one of the largest components of primary biogenic PM worldwide (Hallquist et al., 2009;Seinfeld & Pankow, 2003).
In this study, we investigate the effects of photochemical oxidation on the oxidative potential of biogenic and anthropogenic PM. Samples of each PM type were collected before and after photochemical aging within a laboratory reaction chamber equipped with an ultraviolet lamp. The in vitro alveolar macrophage (AM) assay was used to quantify PM oxidative potential (Landreman et al., 2008;Li et al., 2008;Shafer et al., 2010).

Sampling methods
Photochemical oxidation of primary emissions occurred within a 64-liter stainless steel oxidation flow reactor (OFR) equipped with a single UV lamp (BHK Analamp, emitting radiation at 185 and 254 nm. Upstream of the PM sources, inlet air first passed through an activated carbon denuder and high-efficiency particulate air (HEPA) filter to remove all particles. Within the OFR, a warm, humid environment (22°C/60% RH) was maintained, allowing H 2 O to act as a source of hydroxyl radicals in the UV-catalyzed oxidation reactions, which resulted in SOA formation. While the production of superoxide radicals is possible, the major products of H 2 O oxidation generated in this type of OFR are hydroxyl radicals, especially at less than 80% relative humidity (Seinfeld & Pandis, 2016;Kamens et al., 2011;Jang et al., 2002).
The biogenic sampling setup is depicted in Figure 1. Particlefree inlet air was introduced at a flow rate of 25 lpm. 0.5 lpm of this incoming air stream was diverted into a 250 ml Büchner flask containing a 15-ml glass vial of pure, reagent grade α-pinene. Three small holes in the vial cap allowed for diffusion of α-pinene vapors into the flask. The remaining 24.5 lpm flow of particle-free air proceeded through a humidifier (heated flask containing distilled water) and into the reactor, where it mixed with the α-pinene vapors, resulting in a dilution ratio of 50:1. The anthropogenic sampling setup is depicted in Figure 2. Exhaust from a four-stroke single cylinder gasoline generator (Honda SHX1000, 49cc displacement, 8.0:1 compression ratio, operating at 3000 RPM) was drawn through a rotating disk dilutor (RDD; Testo Engineering, MD19-3E) operating at a dilution ratio of 50:1. 5 lpm of the diluted engine exhaust was diverted into the reaction chamber, where it mixed with 20 lpm of humidified, particle-fee air, resulting in a total dilution ratio of 250:1.
Particles were collected downstream of the reaction chamber on Teflon and quartz filters. In the POA condition, PM was collected as the α-pinene or engine emissions passed through a dark OFR. In the SOA condition, the aerosol stream was sampled while the UV lamp inside the reactor was on. During this condition, the sample stream was diverted for the first 90 minutes to ensure that the reactor was operating under steady-state conditions while samples were collected.

Filter conditioning
Prior to sampling, quartz filters were baked in a furnace oven at 500°C for 5 hours. Teflon filters were conditioned for 24 hours in a controlled environment (23°C and 46% relative humidity) before weighing. Teflon filters were weighed before and after sampling to determine the mass collected with an MT5 Microbalance (Mettler-Toledo Inc., Columbus, OH, USA). Mass collected on quartz filters was calculated based on the aerosol concentration (from Teflon filters) and sampling flow rate. After sampling, all filters were placed in petri dishes lined with baked aluminum foil, sealed with Teflon tape, and stored in a refrigerated environment until analysis.

Laboratory analyses
Quartz filters were analyzed for elemental carbon (EC) and organic carbon (OC) content by the National Institute for Occupational Safety and Health (NIOSH) Thermal Optical Transmission (TOT) Method 5040, using a flame ionization detector (FID) to quantify evolved carbon as CH 4 (Birch & Cary, 1996;Peterson & Richards, 2002).
Each Teflon filter (PM sample or blank) was divided into two sections, one for the macrophage assay and one for chemical analysis. The filter halves used for the cell assay underwent sonication in 900 μL of purified water for approximately 16 hours at room temperature to extract the water-soluble components. These filters were then removed from the aqueous extracts, and 10x concentrated salts-glucose medium (SGM) was added to create buffered PM extract solutions for use in the PM treatments.
In preparation for PM exposures, the non-adherent fraction of macrophage cells was harvested from flasks and concentrated by centrifugation (750 RPM) for 5 minutes. The culture medium was removed, and the cell suspension was diluted to a concentration of 1,000 cells/μL. 100 μL aliquots of these suspended cells were then pipetted into each well of 96-well plates (100,000 cells/well) and incubated at 37°C/5% CO 2 for 2 hours. Following the 2-hour incubation period, nearly all macrophages were adhered to each well bottom. During PM treatments, supernatant media was aspirated and replaced with 100 μL of sample extract or blank control solution in each well.
The macrophage cells were exposed to each type of PM sample for 2.5 hours, using 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) as a fluorescent probe to quantify the cellular formation of oxidative species. The non-fluorescent DCFH-DA acts by entering the cell, where it is de-acetylated by cellular enzymes to yield 2,7-dichlorodihydrofluorescein (DCFH), also non-fluorescent. DCFH is then oxidized by reactive species, generated during the cellular  Zymosan was included as a positive control at a concentration of 0.125 mg-zymosan/mL in a buffered SGM solution.
Results are reported as the increase in fluorescence due to sample treatments relative to the fluorescence observed in the untreated control condition (Landreman et al., 2008;Shafer et al., 2010).

Results
EC/OC results are presented in Figure 3.

Summary and conclusions
The findings of the current reaction chamber study indicate that both anthropogenic and biogenic SOA induce greater cellular oxidative stress than primary engine exhaust. This effect was found to be largest in response to engine exhaust SOA, thus implicating anthropogenic PM as the major contributor to adverse human health effects in urban environments, though the contribution of biogenic SOA can be quite significant in some geographical areas. Atmospheric aging of PM increases its intrinsic oxidative potential many fold, and thus photochemistry in a region that experiences abundant sunshine, long days, and/or stagnation of circulating air due to an inversion layer or some other reason, may increase the toxicity of PM over time.
While the macrophage assay is generally more sensitive to water-soluble components of PM, as compared to non-polar PM species, the large cellular oxidative stress response to SOA, especially anthropogenic SOA, cannot be attributed entirely to hydrophilicity. Biogenic organic compounds are generally more hydrophilic than organic products of anthropogenic processes such as combustion, yet greater ROS activity in response to anthropogenic PM was observed. The alveolar macrophage assay is considered an excellent model of inhalation toxicity, and increased ROS formation reliably indexes greater toxicity, whether ultimately due to increased bioavailability of harmful molecules or simply greater quantities of harmful PM species (Landreman et al., 2008). Thus, the results of this study clearly indicate that oxidized PM species, whether of biogenic or anthropogenic origin, are more toxic that primary PM, regardless of water-solubility, with the greatest toxicity resulting from anthropogenic SOA exposures.

Data availability
The following raw data sets are provided as comma separated values (.csv) files: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. 1.

5.
6. The authors have investigated the oxidative potential of POA and SOA from two different sources namely alpha-pinene and gasoline engine exhaust. The experimental setup included an UV chamber (oxidation flow reactor), to mimic the sun light's UV rays, to compare primary and secondary organic aerosol-induced radical generation under light and in dark. The comparison could contribute great value to the manuscript if additional parameters as listed below are included in it:

Open Peer Review
Page 3: Right column: Line 4: The authors can address why they have selected only Hydroxyl radicals in the investigations. In some experiments, where UV rays are used to excite the organic aerosols, the elicitation of superoxide radicals is also possible.
Page 3: Right column: Lines 28-30: The statement "aerosol stream was sampled while a UV lamp was on, following a 90-minute reaction period." is not clear. Does that mean the whole sample streaming is done for a continuous 90 minutes? Was it the same for the aerosol sampling done in dark OFR?
Page 3: Right column: Line 33: The information of the control sample needs to be included here.
Page 4: Left column: Line 23: The analysis part has some information missing such as incubation time for cell growth, and are the same generation (life cycle) used for analysis?
Page 4: Left column: Line 25: The cell exposure study has some basic information missing -PM dose, route of exposure (directly on filters or on PM extracts), number of times analysed (duplicate or triplicate). Please include for clarity.
Page 5: Figure 4: The biogenic organic compounds (for example: the alpha-pinene) are believed to be more hydrophilic compared to engine exhaust organics. Please include a discussion if the water solubility of samples is also driving the difference in oxidative stress.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed. Text Added: "While the production of superoxide radicals is possible, the major products of H O oxidation generated in this type of OFR are hydroxyl radicals, especially at less than 80% relative humidity (Seinfeld & Pandis, 2016;Kamens et al., 2011;Jang et al., 2002)." : Comment 2 Page 3: Right column: Lines 28-30: The statement "aerosol stream was sampled while a UV lamp was on, following a 90-minute reaction period." is not clear. Does that mean the whole sample streaming is done for a continuous 90 minutes? Was it the same for the aerosol sampling done in dark OFR?
Text Revision: "In the SOA condition, the aerosol stream was sampled while the UV lamp inside the reactor was on. During this condition, the sample stream was diverted for the first 90 minutes to ensure that the reactor was operating under steady-state conditions while samples were collected." : Comment 3 Page 3: Right column: Line 33: The information of the control sample needs to be included here.

Authors' Response:
The only control samples were the filter blanks submitted to the laboratory for chemical and toxicological analyses as discussed in Section 2.3. Discussion of the control condition utilized in the macrophage assay (untreated cells) is included in the response to Comment 5.
: Comment 4 Page 4: Left column: Line 23: The analysis part has some information missing such as incubation time for cell growth, and are the same generation (life cycle) used for analysis?
Text Added: "In preparation for PM exposures, the non-adherent fraction of macrophage cells was harvested from flasks and concentrated by centrifugation (750 RPM) for 5 minutes. The culture medium was removed, and the cell suspension was diluted to a concentration of 1,000 cells/μL. 100 μL aliquots of these suspended cells were then pipetted into each well of 96-well plates (100,000 cells/well) and incubated at 37°C/5% CO for 2 hours. Following the 2-hour incubation period, nearly all macrophages were adhered to each well bottom. During PM treatments, supernatant media was aspirated and replaced with 100 μL of sample extract or blank control solution in each well." : Comment 5 Page 4: Left column: Line 25: The cell exposure study has some basic information missing -PM dose, route of exposure (directly on filters or on PM extracts), number of times analysed (duplicate or triplicate). Please include for clarity.
Text Added: "Each Teflon filter (PM sample or blank) was divided into two sections, one for the macrophage assay and one for chemical analysis. The filter halves used for the cell assay underwent sonication in 900 μL of purified water for approximately 16 hours at room temperature to extract the water-soluble components. These filters were then removed from the aqueous extracts, and 10x concentrated salts-glucose medium (SGM) was added to create buffered PM extract solutions for use in the PM treatments.""Each treatment or control was run in triplicate (3 wells each). Additionally, several untreated and method blank controls were included on each 96-well plate. Zymosan was included as a positive control at a concentration of 0.125 mg-zymosan/mL in a buffered SGM solution. Results are reported as the increase in fluorescence due to sample treatments relative to the fluorescence observed in the untreated control condition." : Comment 6 Page 5: Figure 4: The biogenic organic compounds (for example: the alpha-pinene) are believed to be more hydrophilic compared to engine exhaust organics. Please include a discussion if the water solubility of samples is also driving the difference in oxidative stress.

Text Added (to Summary and Conclusions):
"While the macrophage assay is generally more sensitive to water-soluble components of PM, as compared to non-polar PM species, the large cellular oxidative stress response to SOA, especially anthropogenic SOA, cannot be attributed entirely to hydrophilicity. Biogenic organic compounds are generally more hydrophilic than organic products of anthropogenic processes such as combustion, yet greater ROS activity in response to anthropogenic PM was observed. The alveolar macrophage assay is considered an excellent model of inhalation toxicity, and increased ROS formation reliably indexes greater toxicity, whether ultimately due to increased bioavailability of 2 formation reliably indexes greater toxicity, whether ultimately due to increased bioavailability of harmful molecules or simply greater quantities of harmful PM species (Landreman et al., 2008). Thus, the results of this study clearly indicate that oxidized PM species, whether of biogenic or anthropogenic origin, are more toxic that primary PM, regardless of water-solubility, with the greatest toxicity resulting from anthropogenic SOA exposures." The work of Lovett presents interesting data on the possible inflammatory effects of SOA from traffic et al. as well as biogenic (pinene) origin. The method setup is well designed, although the variables, such as conditions of relative humidity and temperature, could have been studied in different values. On the other hand, the results show that the biogenic SOA have similar high inflammatory effect in the test compared to traffic SOA, which is an important fact, and indicates that effects have been observed in real-world data. However, the study could have been more complete and higher quality if the researchers would have made an effort to detect and quantify the molecular chemical compounds that are present in the SOA fractions.
The work is suitable for indexing.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. it is of an acceptable scientific standard.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative Your article is indexed in PubMed after passing peer review Dedicated customer support at every stage For pre-submission enquiries, contact research@f1000.com