Titanium (Ti), which has an atomic number 22 and atomic weight 47.88, is a transition element belonging to Group 4 (IVB) in the middle of the Periodical Table. It is the ninth most abundant element in the earth's crust and makes up about 0.25% by moles and 0.57% by weight of the crust (Buettner and Valentine, 2012). Ti is the second most abundant transition metal, after iron (Fe), and the elemental abundance of Ti is about 5 times less than Fe and 100 times greater than copper (Cu). Ti exhibits oxidation states of Ti²⁺, Ti³⁺ (titanous), and Ti⁴⁺ (titanic), of which Ti²⁺ and Ti³⁺ are unstable, while Ti⁴⁺ is the most stable ion. The most important compound is TiO₂, which is mainly used in paints. TiCl₄ is water soluble but is highly volatile and forms spectacular opaque clouds upon contact with humid air. Ti ascorbate is a synthesized compound which is soluble in water and stable up to pH 8.0.

The mineral sources of Ti include anatase, rutile, and brookite, each encompassing about 95% TiO₂ as well as ilmenite (FeOTiO₃) comprising 40–65% TiO₂ and leucoxene (Fe₂O₃ nTiO₃) containing more than 65% TiO₂ (Zhang et al., 2011). These minerals are generally not soluble; thus Ti has been conventionally considered to be inert in the environment. Increasing evidence in the literature, however, suggests that Ti is mobile in rocks under weathering conditions (Kaup and Carter, 1987; Du et al., 2012). Ti may be mobile at the centimeter scale as well as at the profile scale under strong tropical weathering conditions (Cornu et al., 1999). Higher Ti contents occur in tropical soils, particularly in lateritic soils and laterites, such as 15% in Hawaii soils (Sherman, 1952); 15% in Norfolk Island soils (Hutton and Stephens, 1956), and 3.4% in Australian soils (Stace et al., 1968). Ti in surface soils worldwide ranges from 0.02 to 2.4% with a mean of 0.33%; Ti in soil solutions is about 30 mg L⁻¹ (Kabata-Pendias and Mukherjee, 2007). Ti in river waters ranges from 0.02 to 2.3 μg L⁻¹, and the worldwide average is estimated to be 0.49 μg L⁻¹ (Kabata-Pendias and Pendias, 2001). Drinking waters in the US contain Ti from 0.5 to 15 μg L⁻¹ (Anke and Seifert, 2004). Ti also exists in the atmosphere with global median values of 7 ng m⁻³ in the remote regions (away from anthropogenic releases) and 85 ng m⁻³ in polluted zones. Ti concentrations in the air of the US vary from 10 to 100 ng m⁻³ and can increase up to ≤1,000 ng m⁻³ in industrial regions (Kabata-Pendias and Mukherjee, 2007).

Titanium dioxide nanoparticles (TiO₂NPs) are another form of Ti in the environment. TiO₂NPs are produced worldwide at an estimated 88,000 t per year (Keller et al., 2013) and are utilized widely in the cosmetic, food, painting, and plastic industries. Due to their photoprotective and photocatalytic roles, TiO₂NPs are also used for plant protection and environmental remediation. It is estimated that the concentrations of TiO₂NPs in soils could reach 0.13 μg kg⁻¹ yr⁻¹ in Europe, and TiO₂NPs in soils amended with sewage could be much higher up to 1,200 μg kg⁻¹ yr⁻¹ (Sun et al., 2014). With the increased exploration of nanomaterials for novel commercial applications, TiO₂NPs in soils could increase from 3 to more than 5,000 μg kg⁻¹ yr⁻¹ (Gogos et al., 2012; Kah et al., 2013).

The earth contains 92 elements, of which 82 can be found in plants (Reimann et al., 2001). Ti contents in plants range from 1 to 578 mg kg⁻¹ with a mean of 33.4 mg kg⁻¹ across the listed species (Table 1) excluding two Ti accumulators: horsetail (Equisetum spp.) and beach morning glory [Ipomoea pes-caprae (L.) R. Br.]. There are several factors affecting plant absorption of Ti: (1) Plant species differ in Ti uptake. Ti concentrations vary from 20 mg kg⁻¹ in red cabbage (Brassica oleracea var. capitata f. rubra) to 1,900 mg kg⁻¹ in the wood of pedunculate oak (Quercus robur L.) (Dumon and Ernst, 1988). Ti in horsetail ranged from 42 to 14,000 mg kg⁻¹ when grown in soils rich in lead and zinc (Cannon et al., 1968). (2) Plants respond to Ti addition regardless of soil application or hydroponic culture. Increased Ti application elevates Ti concentrations in crops, such as cabbage (Hara et al., 1976), common bean (Phaseolus vulgaris L.) (Ram et al., 1983), corn (Zea mays L.) (Pais, 1983), and pepper (Capsicum annuum L.) (Giménez et al., 1990). Plant roots accumulate more Ti with a small amount transported to shoots (Kelemen et al., 1993). (3) Soil pH significantly affects the absorption of Ti in plants. Acid sandy soil (pH 3.1) increased Ti solubility resulting in Ti concentrations in leaves of gray hair grass (Corynephorus canescens P. Beauv.) and Sheep's sorrel (Rumex acetosella L.) up to 142 and 207 mg kg⁻¹, respectively; however, leaf Ti concentrations of the same species were only 2.4 and 4.8 mg kg⁻¹, respectively when grown in a soil with nearly identical total Ti concentrations but a pH at 4.9 (Ernst, 1985). Ti concentration in beach morning glory ranged from 310 to 480 mg kg⁻¹ when grown in the ilmenite soil with a pH range of 7.8–7.9, whereas Ti concentration was 910 to 1,300 mg kg⁻¹ in a pH range from 7.3 to 7.4 (Ramakrishna et al., 1989). (4) Foliar application is more effective for Ti absorption. Ti content in leaves and stems increased with Ti sprays but the increase was limited in soil application (Wojcik and Wojcik, 2001). Tapertip hawksbeard (Crepis acuminata Nutt.) is a dust-indicator plant, and seedlings of this species showed an 11-fold increase in Ti after being exposed to contaminated soil dusts (Cook et al., 2009). (5) Ti deficiency symptoms have not been described in plants. Ti supplied at low concentrations has been shown to positively affect plant growth (Figure 1) but causes phytotoxicity at high concentrations (Wallace et al., 1977).

The biological role of Ti in plants has been studied for more than 100 years. Pellet and Fribourg (1905) were the first to study Ti in soils and sugar cane (Saccharum spp.) and sugar beets (Beta vulgaris L.). Traetta-Mosca (1913) observed that Ti enhanced the growth of tobacco (Nicotiana tabacum L.) leaves and believed that Ti was an inherent constituent of the ash from all plants. They proposed that Ti might participate in plant metabolism as a redox catalyst. Geilmann (1920) found that Ti mainly accumulated in assimilation organs. A systematic study of plant responses to different concentrations of Ti by Němec and Káš (1923) showed that optimal levels of Ti caused increased plant growth and development and increased the intensity of green color (higher chlorophyll content) of mustard (Brassica arvensis L.), pea (Pisum sativum L.), and alfalfa (Medicago sativa L.). Subsequently, a great deal of attention from the 1920s to early 1970s has been focused on the analysis of Ti contents in wild and cultivated plants (Dumon and Ernst, 1988). Pais et al. (1977) synthesized a Ti compound called Ti-ascorbate with a trade name of Titavit. It was produced by chelating TiCl₄ with ascorbic acid in the presence of gaseous HCl. Ti-ascorbate is water soluble, stable up to pH 8, and also not toxic to animals. Since then, Ti-ascorbate has been widely used for Ti-related plant experiments (Pais, 1983; Carvajal and Alcaraz, 1998; Hrubý et al., 2002; Cigler et al., 2010). A commercial product called Tytanit® containing 5% MgO, 10% SO₃, and 0.85% other titanium complex was developed and used in central and eastern European countries for improving crop production. Ti has also been used as a beneficial element in China for crop production (Li et al., 2011).

Several explanations have been proposed concerning the actions of Ti as a beneficial element to plants, including (1) participation in N fixation in the nodules of legumes (Konishi and Tsuge, 1936); (2) influence on plant metabolism by increasing absorption of other nutrient elements, such as Fe and Mg (Dumon and Ernst, 1988; Simon et al., 1988); (3) involvement in redox system reactions (Ti⁴⁺/Ti³⁺ with Fe³⁺/Fe²⁺) thus improving the Fe activity in plant tissues (Carvajal et al., 1995) or interaction with Fe in electron transport chain and decrease of the photosystem II efficiency at a high Ti concentration (Cigler et al., 2010); (4) stimulation of enzymatic activities and photosynthesis (Carvajal and Alcaraz, 1998); and (5) hormesis (Hrubý et al., 2002; Kuzel et al., 2003). Among these claims, Ti participation in N fixation has not been documented thereafter the initial report (Konishi and Tsuge, 1936); as a result, this claim may not be valid. Hormesis is a term used by toxicologists to refer to a biphasic dose response to an environmental agent characterized by low dose stimulation or beneficial effects and a high dose inhibitory or toxic effect (Mattson, 2008). It is a biological phenomenon for almost any chemical element or drug in living things, and it cannot be considered a specific mechanism for Ti actions in plants. The other explanations are mainly focused on the physiological roles of Ti in plants and have not explored any cellular or molecular mechanisms underpinning its actions.

A common characteristic of beneficial elements is their ability to positively interact with one or more essential elements, primarily by partial substitution of essential elements: such as sodium (Na) with potassium (K), selenium (Se) with sulfur (S), cobalt (Co) with nickel (Ni), and silicon (Si) with boron (B), manganese (Mn), and phosphorus (P). Such interactions could be synergistic at a certain concentration range but may become antagonistic when the concentration is too high. For example, when K supply becomes limited in soils, Na can partially substitute for K in osmoregulation (Marschner, 2011). Both elements are alkali metals in the Group 1 column of the Periodic Table and have similar physical and chemical properties. Like K, Na can enter plant cells through K channels (Demidchik et al., 2002). Se and S are both Group VIA elements in the Periodic Table and share similar chemical properties. Se is absorbed by plants in the form of selenate through sulfate transporters (Cabannes et al., 2011). S uptake is enhanced by rhizosphere selenate; however, Se toxicity occurs if Se and S compete for a biochemical process (White et al., 2004). Co and Ni are both transition metals and are generally found together in nature. Co is synergistically related to Ni, and reports showed that toxic Co levels of 10–20 mg kg⁻¹ dry mater were associated with excess Ni (Anderson et al., 1973). This is because Co and Ni share the same plasma membrane carriers (Pilon-Smits et al., 2009).

We here propose that the beneficial roles Ti plays in plants lie in its interaction with other nutrient elements, primarily Fe. This proposal is not new and has been postulated by Simon et al. (1988), Carvajal and Alcaraz (1998), and Cigler et al. (2010). More specifically, we hypothesize that Ti and Fe have synergistic and antagonistic relationships. When plants encounter Fe deficiency, Ti could induce the expression of genes related to Fe acquisition, enhancing Fe uptake and utilization and subsequently improving plant growth. Plants could have proteins that either specifically or nonspecifically bind with Ti. When Ti concentration is high in plants, it may compete with Fe for ligands or proteins. The competition could be severe, resulting in Ti phytotoxicity. As such, the beneficial effects of Ti could be particularly visible or measurable during the time when plants are near to or are experiencing Fe deficiency. This hypothesis relies on the beneficial effects of Ti that have been reviewed above and will be elaborated further in subsequent sections of this review.

Ti and Fe have similar physical and chemical properties. Both Ti and Fe are transition metals. The ionic radius and Pauling electronegativity of Ti are 0.7 Å and 1.54; the same parameters for Fe are 0.9 to 0.7 Å and 1.83. Ti and Fe occur together in nature. During magmatic processes, Ti follows Fe in magmatic crystallization. Ti⁴⁺ is predominantly partitioned into Fe-Ti or Fe oxides, such as ilmenite (FeTiO₃) and magnetite (Fe₃O₄), or into one or more of the TiO₂ phases, rutile (TiO₂), and anatase (TiO₂). The Ti-Fe-oxides and their relationships have been illustrated by triangular FeO-TiO₂-Fe₂O₃ diagrams (Bowles et al., 2011). Ilmenite (FeTiO₃) is the most widespread form of TiO₂-bearing mineral around the world, and it provides 90% of the total world Ti. Ti has been shown to be mobile in rocks under weathering conditions and also in soils (Cornu et al., 1999). It could be possible that adaptation of plants to soils containing heavy mineral sands (derived from the weathering of ilmenite) might enable roots to absorb both Fe and Ti. Due to the abundance of Fe in soil relative to Ti and its biological functionality, more Fe is absorbed by and translocated in plants. As a result, Fe has been fulfilling much more important roles in plants. Fe is thus considered an essential element to plants, while Ti plays a complementary role, i.e., it is often found along with Fe and plays both synergistic and antagonistic roles depending on Fe concentrations in plant cells.

Plant uptake of ions through roots or leaves involves both passive absorption and active transport. Passive absorption is facilitated by concentration gradients of an ion, while active transport is driven by the electrochemical gradient generated by H⁺-ATPase to allow selective ions to move across the plasma membrane through specific carriers or transporters.

Ti absorbed via roots or leaves is translocated to the other organs. Like most transition elements, root-absorbed Ti is largely accumulated in the roots with a small amount transported to shoots through xylem stream (Kelemen et al., 1993). Ti absorbed by leaves is translocated via phloem flow.

Plant cells contain numerous iron-containing proteins which can be mainly classified into three groups: iron-sulfur cluster proteins, hemeproteins, and non-heme/non-Fe-S proteins (Zhang, 2015). These proteins use Fe as a cofactor and perform critical roles in photosynthesis, genome stability, electron transfer, and oxidation-reduction reactions. Plants have evolved sophisticated mechanisms to maintain iron homeostasis for the assembly of functional iron-containing proteins, thereby ensuring genome stability, cell development, electron transport chain of photosynthesis and respiration in chloroplasts and mitochondria, respectively (Kobayashi and Nishizawa, 2012). Fe is also essential for reactive oxygen species (ROS) detoxification, chlorophyll biosynthesis, period length control of circadian rhythm, and activity of numerous metal-dependent enzymes (Alscher et al., 2002; Moseley et al., 2002; Chen et al., 2013). Most of the Fe in leaves is found within the chloroplasts where photosynthesis takes place to assimilate C and produce O₂. In addition to the general mitochondrial Fe-S cluster synthesis pathway, chloroplasts are autonomous for their Fe-S cluster synthesis (Zhang, 2015). It is within this plant specific subcellular compartment that ferritins store and buffer Fe, thereby participating in remediating oxidative stress. Ferritins are plastid proteins whose abundance is strictly controlled at a transcriptional level by the Fe status of the cells (Kobayashi and Nishizawa, 2012; Zhang, 2015).

In the case of Fe and Ti interactions, Ti effects could become more pronounced when plants had deficient supply of Fe. Under such conditions, application of Ti could induce the expression of IRT in nongraminaceous tobacco plants and YS1 in graminaceous corn plants. The expression of ferritin genes could also be enhanced by Ti application. The induced expression of these genes under limited Fe supply might suggest that some roles Ti would play could be the maintenance of Fe homeostasis at the cellular level, thus improving plant growth. Carvajal and Alcaraz (1995) demonstrated that foliar application of Ti-ascorbate resulted in an increase of Fe concentrations in leaves, fruits, chloroplasts, and chromoplasts of red pepper plants. Foliar application of Ti resulted in 39% and 35.7% increase of Fe in peel and flesh of peach fruit (Alcaraz-Lopez et al., 2004a,b). Leaves of paprika pepper sprayed with Ti-ascorbate increased Fe uptake by 50% in a greenhouse experiment and close to 100% in a field experiment, and leaf peroxidase and catalase activities also significantly increased due to the Ti-ascorbate application (Carvajal et al., 1995). These results provide further evidence supporting our hypothesis that the synergetic roles Ti plays become more noticeable when plants encounter low Fe supply. Under a limited Fe supply, application of an appropriate concentration of Ti would induce IRT or YS1 expression, thus enhancing Fe uptake. Increased Fe uptake would increase chlorophyll biosynthesis, subsequently increasing net photosynthesis. Increased photosynthesis directly couples with NO3− assimilation in chloroplasts, which is known as nitrate photoassimilation (Searles and Bloom, 2003). The increased photosynthesis would enhance the expression of nitrate transporter genes, consequently increasing N uptake. The increased uptake of NO3− could improve plant growth and in turn enhance absorption of other ions. For example, a 7-fold increase in N uptake by rhododendron (Rhododendron spp. cv. P.J.M. Compact) was associated with a 3 to 4-fold increase in the uptake rate of phosphorus, potassium, and sulfur, and ~2-fold increase in the uptake rate of magnesium and calcium (Scagel et al., 2008). Additionally, IRT1 belongs to the ZRT/IRT-like protein (ZIP) gene family, which plays a major role in Fe/Zn (zinc) uptake (Guerinot, 2000). IRT1 can also transport Zn, Co, Mn, and cadmium (Cd) (Eide et al., 1996; Connolly et al., 2002; Varotto et al., 2002; Vert et al., 2002). YS1 functions as a proton-coupled symporter for various DMA-bound metals, including Fe³⁺, Zn²⁺, Cu²⁺, and Ni²⁺ (Kakei et al., 2012). This may explain why the application of Ti also increases plant uptake of other nutrient elements.

Ti may act antagonistically with Fe resulting in Ti toxicity in plants. If Ti concentration is too high, it could interfere with biological roles of Fe, resulting in Ti toxicity. Cigler et al. (2010) measured chlorophyll fluorescence of spinach plants after treatment by a combination of Fe and Ti. They found that Ti at a high level affects Fe-containing proteins in electron transport, primarily the PSI, slowing down the PSII efficiency. If Ti and Fe were equally present in the medium, the Ti impact on the PSI was lowered, probably due to competition for binding sites.

Ti in both bulk and nanoparticle forms has been used for suppressing crop diseases (Paret et al., 2013; Servin et al., 2015). Chao and Choi (2005) reported that severity and incidence of curvularia leaf spot [Curvularia lunata (Wakker) Boedijn] and bacterial leaf blight (Xanthomonas oryzae pv. oryzae) in cereal crops were reduced with TiO₂ application. Similar results were observed on field-grown cowpea (Vigna unguiculata Walp.) where cercospora leaf spots caused by Cercospora rosicola Pass. and brown blotch caused by Mycosphaerella cruenta Sacc. were significantly suppressed by application of TiO₂ (Owolade and Ogunleti, 2008). TiO₂ has been shown to control bacterial leaf spot (Xanthomonas hortorum pv. pelargonii) on geranium (Pelargonium x hortorum L.H. Bariley) and (Xanthomonas axonopodis pv. poinsettiicola) on poinsettia (Euphorbia pulcherrima Willd. Ex klotzsch.) (Norman and Chen, 2011). Additionally, the use of TiO₂ in recycled irrigation water was shown to eliminate both fungal and bacterial pathogens (Yao et al., 2007).

The antimicrobial roles of TiO₂ are related to the oxidation processes even though the role of Ti-uptake resultant biological activities could not be ruled out. Recently, the photocatalytic process by UV/TiO₂ is receiving increased attention due to the low cost and relatively high chemical stability of TiO₂, especially in aqueous environments. It generates singlet oxygen and superoxide anion which both cause damaging cellular oxidation. Therefore, TiO₂ has been used for controlling some bacterial and fungal pathogens in crop production (Yao et al., 2007; Owolade and Ogunleti, 2008; Norman and Chen, 2011) and also for decontaminating toxic organic pollutants in water treatment (Lazar et al., 2012). TiO₂NPs have been shown to degrade organic pesticides and herbicides in soils via redox reactions, photocatalysis, and thermal destruction under irradiation (Mir et al., 2014; Li et al., 2016). Photocatalytic TiO₂ has been used to kill cancer cells in human (Thevenot et al., 2008), and biomedical applications of TiO₂NPs are promising and could play important roles for improving health care, especially cancer treatment (Yin et al., 2013).

Evidence accumulated over the last 100 years suggests that Ti is relatively mobile in soils, occurs in soil solution, and is available to plants. Plants are able to absorb Ti through either roots or leaves, and Ti concentrations in plant tissues are either equal to or higher than some essential nutrient elements. Ti has been shown to improve plant performance at low concentrations. In the present article, we propose Ti and Fe have synergistic and antagonistic relationships. Ti may induce the expression of genes related to Fe acquisition, enhancing Fe uptake and utilization when plants encounter Fe deficiency. The interaction of plants with Ti as well as with Fe may result in the occurrence of Ti binding proteins in plants that either specifically bind with Ti or nonspecifically share with Fe or other elements. When Ti levels are high in plants, Ti may cause phytotoxicity. This hypothesis is not new but is updated based on the current available information. With the advances in omics technologies, we anticipate that this hypothesis will be tested and improved.

All authors contributed to the acquisition and interpretation of available literature and the conception of the work. JC, SL, and XYW wrote the manuscript, and all authors revised the manuscript and approved this final version.