**4. Discussion**

Carniolan honey bees exposed to HMF su ffer increased death, especially after feeding on it for 15 to 30 days. Bee survival, therefore, could be improved by reducing worker exposure to HMF in their foodstu ffs. In a large-scale experiment, a relatively low concentration of HMF (150 mg/kg) found in acid-hydrolyzed inverted sugar syrup induced 50% bee mortality within 16 days after the start of feeding. Likewise, high-fructose corn syrup, a saccharose replacement for honey bees, containing 150 mg/kg HMF also induces 50% of bee mortality within about 19 days of feeding [14]. An HMF concentration of 30 mg/kg to 48 mg/kg is supposed to be harmless to honey bee workers, whether or not bees are overwintering [11]. The concentration of HMF in winter food stores is lower than that of initial syrups deposited by bees. Perhaps honey bees can safely metabolize or detoxify small quantities of HMF in their stored foods [25]. Except for these small-scale studies, few studies have systematically examined the e ffect of HMF on bee health and physiology, which is surprising given that concentrations of 7500 ppm HMF (mg/kg) or higher have been reported to cause massive bee kills, i.e., 100% mortality of bee larvae with LC50 for larvae ranging from 2424 ppm to 4280 ppm. Thus, larvae appear far more sensitive to HMF than adult honey bees [26]. Clearly, our results coupled with evidence cited earlier confirm the toxicity of HMF to adult and developing honey bees and the need for its mitigation in commercial bee yards. A deadly chemical interaction may occur when HMF builds up in particular foodstu ffs. For instance, sugar syrup or high-fructose corn syrup enhance the acute toxicity of HMF and double rates of honey bee mortality (from 40% in Apifonda controls to 80% in sugar syrups) at a high concentration of around 1500 mg/kg of HMF when compared with its combination with Apifonda candy. The e ffects of HMF in the first two weeks of feeding are sublethal for bees, but after 14 days, HMF starts to become fatal. More studies are needed to assess the tolerances of honey bees to prolonged oral exposure to HMF.

The mechanism through which HMF sickens or kills honey bees is unknown. We surmise that the midgut is the first bodily tissue exposed to the activity of HMF as well as to other toxins, including pathogens. Midgut epithelial cells are the first line of defense against both pathogens

and toxins [27]; [28] and potential dysbiosis together with other disorders can present an important source of energy dysregulation [29] in epithelial midgut cells. Our midgut analysis of worker honey bees demonstrates hypertrophic enlargement of the digestive cells in the first 5 days after HMF treatment. Later, ~10 days post-HMF treatment, numerous a ffected cells were released into the lumen, as evidenced by observable apoptotic cell death in the apical region of midgut villi. Between the 10th and 15th day of feeding, 1000 mg/kg of HMF and 1500 mg/kg of HMF increased the cell death rate, resulting in the shedding of dead cells from the epithelium into the midgut lumen. In contrast, 500 mg/kg of HMF resulted in low levels of cell death.

Initially, such changes to the apical epithelium and some cell death are adaptive, but uncontrolled epithelial cell losses without adequate regeneration leads to loss in midgut function [30,31]. In a control group where bees received only sugar candy (Apifonda), high proportions of apoptotic columnar cells were found, which indicates a high level of cell turnover. In bees exposed to high doses of HMF, normal apoptosis was followed by morphological changes typical of necrotic deletion. In this study, we found that caged bees were sensitive to the HMF treatment, as observed through changes at the tissue level, and its potential detrimental e ffect can result in higher bee mortality.

In addition to the di fferences in the mortality rates induced by HMF treatment, there was accompanying histopathology whereby lesions formed within midgut cells. HMF induced a higher cell death rate in comparison with untreated bees by the 5th, 10th, and 15th day post-treatment. Early and accelerated levels of apoptosis induced by HMF toxicity may serve as a possible defense mechanism in the midgut that impedes HMF from a ffecting neighboring epithelial cells, a similar mechanism to that employed when cells are infected with pathogenic organisms [32]. The midgut of honey bees also undergoes hypertrophic changes in the first day after treatment and continued to slough dead cells into the lumen. Later, a reduction in apoptotic cell deletion and cell lysis occur along the entire midgut epithelium. Thus, damaged epithelium regenerates as dying cells gradually separate from the basal lamina. The new epithelium might also form just after the complete discharging of the degenerated epithelium, when only regenerative cell groups and remnants of cell membranes are observed [33]. This regenerative process in the midgut may importantly enhance bee survival after prolonged exposure to HMF.

Clearly, HMF has a dosage-dependent cytotoxic e ffect on honey bee digestion; both sublethal and subclinical changes to the midgut occur at the cellular level before bees eventually die from high doses. Adverse e ffects of HMF feeding needs study at the colony level because bees may rely on unknown behavioral mechanisms to mitigate the toxic e ffects of HMF-contaminated food stores on bee and brood survival.

**Author Contributions:** Conceptualization, A.G. and S.J.; methodology, A.G.; investigation, S.J. and A.G.; data curation, B.S. and A.G., writing—original draft preparation, A.G. and S.J., writing—review and editing, B.S.; supervision, A.G. and B.S. Allauthors have read and agreed to the published version of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Slovenian Research Agency, Research core funding No. P1-164; Research for improvement of safe food and health.

**Acknowledgments:** We are grateful for the assistance provided by Mitja Nakrst and Maja Ivana Smodiš Škerl in the experiments. The study was supported and research topics discussed by COLOSS (Prevention of honey bee COlony LOSSes, http://coloss.org/).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
